fix(desktop): don't start the system tray on macOS (#3223) (#3225)

Disable the system tray on macOS: startTray() was called unconditionally at startup and fyne.io/systray creates its NSStatusItem (an NSWindow) off the main thread, crashing every macOS launch of 1.2.0. traySupported() now returns false on darwin; close-to-background there already uses runtime.Hide + the Dock.

Closes #3223

.env.example

@@ -1,10 +1,5 @@
-DEEPSEEK_API_KEY=sk-your-key-here
-DEEPSEEK_BASE_URL=https://api.deepseek.com
-# DEEPSEEK_API_BASE_URL is accepted as an alias of DEEPSEEK_BASE_URL.
-REASONIX_LOG_LEVEL=INFO
-REASONIX_TRANSCRIPT_DIR=./transcripts
+# Copy to .env and fill in. Only secrets belong here — a provider's base_url and
+# model are configured in reasonix.toml, where api_key_env points at these variables.
 
-# Optional Telegram remote channel
-# TELEGRAM_BOT_TOKEN=123456:botfather-token
-# TELEGRAM_OWNER_USER_ID=123456789
-# TELEGRAM_ALLOWLIST=123456789,987654321
+DEEPSEEK_API_KEY=
+MIMO_API_KEY=

.gitattributes

@@ -1,11 +1,2 @@
-# Force LF line endings on every checkout, regardless of platform.
-# Biome's formatter is strict about CRLF — without this, a Windows clone
-# with autocrlf=true will land with CRLF in the working tree even though
-# the repo stores LF, and `npm run lint` (and prepublishOnly with it)
-# fails. See docs/style.md if you ever add binary types here.
-* text=auto eol=lf
-
-# Lockfiles + the codemirror bundle are generated; treat them as
-# auto-detected text but never coerce.
-package-lock.json text eol=lf
-dashboard/codemirror.js text eol=lf
+.githooks/** text eol=lf
+benchmarks/**/*.sh text eol=lf

.githooks/pre-push

@@ -0,0 +1,5 @@
+#!/bin/sh
+# Go-native pre-push gate replacing the retired v1 npm hook (install: make hooks).
+# Full tests run in CI; vet is the fast, cross-platform local check.
+set -e
+go vet ./...

.github/CODEOWNERS

@@ -0,0 +1,3 @@
+# Every pull request requests review from the maintainers below.
+# Either maintainer can review and merge any change.
+* @SivanCola @esengine

.github/FUNDING.yml

@@ -1,3 +0,0 @@
-github: esengine
-custom:
-  - https://paypal.me/yuhuahui

.github/ISSUE_TEMPLATE/bug_report.md

@@ -1,30 +0,0 @@
----
-name: Bug report
-about: Something is broken
-labels: bug
----
-
-> **Screen flicker, garbled output, leftover artifacts, cursor jumping?**
-> Use the **Display / rendering issue** template instead — it asks for the
-> terminal-specific info we need to diagnose those.
-
-**What happened**
-A clear and concise description.
-
-**Expected**
-What you expected to happen.
-
-**Reproduction**
-Steps or minimal code that reproduces it.
-
-**Environment**
-- Reasonix version (`reasonix --version`):
-- Node version (`node --version`):
-- OS (Windows 11 / macOS 14 / Ubuntu 24.04 / …):
-- Shell (bash, zsh, fish, PowerShell 7, PowerShell 5.1, cmd, …):
-- Terminal app (Windows Terminal, iTerm2, Alacritty, kitty, WezTerm, **VSCode integrated**, **Cursor integrated**, Hyper, …):
-- DeepSeek model (e.g. `deepseek-v4-flash`, `deepseek-v4-pro`):
-
-**Logs / transcript**
-If using the CLI, attach the relevant chunk of `--transcript`, or run
-`reasonix doctor` and paste the output.

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,60 @@
+name: Bug report
+description: Report a bug in DeepSeek-Reasonix
+title: "[Bug]: "
+labels: ["bug"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for filing a bug. Pick which version line you're on below — it gets
+        labeled (`v1`/`v2`) automatically, no maintainer action needed.
+  - type: dropdown
+    id: version-line
+    attributes:
+      label: Version line
+      description: Which line are you running?
+      options:
+        - "v2 — Go rewrite (1.x), main-v2 (active development)"
+        - "v1 — Legacy TypeScript (0.x), maintenance only"
+    validations:
+      required: true
+  - type: input
+    id: version
+    attributes:
+      label: Exact version
+      description: Output of `reasonix --version`.
+      placeholder: e.g. 1.0.0
+    validations:
+      required: true
+  - type: textarea
+    id: what-happened
+    attributes:
+      label: What happened?
+      description: A clear description of the bug, and what you expected instead.
+    validations:
+      required: true
+  - type: textarea
+    id: repro
+    attributes:
+      label: Steps to reproduce
+      placeholder: |
+        1. ...
+        2. ...
+        3. ...
+    validations:
+      required: true
+  - type: input
+    id: os
+    attributes:
+      label: OS / platform
+      placeholder: e.g. macOS 15.3 (arm64), Ubuntu 24.04, Windows 11
+    validations:
+      required: true
+  - type: textarea
+    id: logs
+    attributes:
+      label: Relevant logs or output
+      description: Paste any error output. Automatically formatted as code.
+      render: shell
+    validations:
+      required: false

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,5 @@
+blank_issues_enabled: false
+contact_links:
+  - name: Questions & discussions
+    url: https://github.com/esengine/DeepSeek-Reasonix/discussions
+    about: For usage questions, ideas, and general discussion — please use Discussions instead of opening an issue.

.github/ISSUE_TEMPLATE/display_issue.md

@@ -1,109 +0,0 @@
----
-name: Display / rendering issue
-about: Screen flicker, garbled output, leftover artifacts, cursor jumping
-labels: bug, rendering
----
-
-> Display problems almost always come from the **terminal emulator**, not
-> the shell. Please fill the terminal section carefully — `bash vs PowerShell`
-> tells us very little; `VSCode integrated terminal vs Windows Terminal`
-> tells us everything.
-
-**Symptom** (tick all that apply)
-- [ ] Whole screen flickers / flashes during streaming response
-- [ ] Lines tear or only half-redraw
-- [ ] Stale output left behind after a frame updates
-- [ ] Cursor jumps to wrong column or vanishes
-- [ ] Mojibake / wrong-width characters (e.g. `□`, half-width emoji)
-- [ ] Other (describe below)
-
-**When it happens**
-- [ ] During assistant streaming (token-by-token output)
-- [ ] When tool cards expand / collapse
-- [ ] During scroll-up / scrollback
-- [ ] On terminal resize
-- [ ] On launch / on quit
-- [ ] Other (describe below)
-
-**Terminal — the important part**
-
-Where exactly are you running `reasonix`?
-
-- [ ] **VSCode** integrated terminal — VSCode version: `?`
-- [ ] **Cursor** integrated terminal — Cursor version: `?`
-- [ ] **Windows Terminal** — version: `?`
-- [ ] **cmd.exe** (legacy console host)
-- [ ] **PowerShell ISE** (note: ISE doesn't support ANSI — most things will look broken)
-- [ ] **iTerm2** / **Terminal.app** / **Alacritty** / **kitty** / **WezTerm** / **Hyper**
-- [ ] tmux / screen / mosh — and inside which outer terminal? `?`
-- [ ] Other:
-
-> 💡 **How to find your VSCode / Cursor version**
-> `Help → About` (Windows/Linux) or `Code → About Visual Studio Code` (macOS).
-> Paste the whole panel — version + commit + Electron + xterm.js if shown.
-
-**Diagnostic dump — copy/paste output**
-
-Run **one** of the snippets below in the same terminal where you saw the
-issue, and paste the output here:
-
-<details><summary>PowerShell (Windows)</summary>
-
-```powershell
-reasonix --version; node --version
-$PSVersionTable.PSVersion.ToString()
-[System.Environment]::OSVersion.VersionString
-"TERM=$env:TERM"
-"TERM_PROGRAM=$env:TERM_PROGRAM"
-"TERM_PROGRAM_VERSION=$env:TERM_PROGRAM_VERSION"
-"COLORTERM=$env:COLORTERM"
-"WT_SESSION=$env:WT_SESSION"
-"VSCODE_INJECTION=$env:VSCODE_INJECTION"
-"WSL_DISTRO_NAME=$env:WSL_DISTRO_NAME"
-```
-
-</details>
-
-<details><summary>bash / zsh (macOS / Linux / WSL / Git Bash)</summary>
-
-```bash
-reasonix --version; node --version
-uname -a
-echo "TERM=$TERM"
-echo "TERM_PROGRAM=$TERM_PROGRAM"
-echo "TERM_PROGRAM_VERSION=$TERM_PROGRAM_VERSION"
-echo "COLORTERM=$COLORTERM"
-echo "WT_SESSION=$WT_SESSION"
-echo "VSCODE_INJECTION=$VSCODE_INJECTION"
-echo "WSL_DISTRO_NAME=$WSL_DISTRO_NAME"
-```
-
-</details>
-
-```
-<paste output here>
-```
-
-**VSCode / Cursor users only — terminal settings**
-
-Open Settings (`Ctrl+,`), search `terminal.integrated.gpuAcceleration`,
-report current value: `auto` / `on` / `canvas` / `off` — `?`
-
-Already tried any of:
-- [ ] Switching `gpuAcceleration` to a different value
-- [ ] Detaching the terminal (drag tab into its own window)
-- [ ] Running the same command in a non-VSCode terminal — did it still flicker? `yes / no`
-
-**Reproduction**
-
-Steps that reliably trigger it (commands run, files edited, was a tool
-streaming a long response, was the window being resized, …):
-
-1.
-2.
-3.
-
-**Screen recording (strongly preferred)**
-
-A 5–10s GIF or MP4 is worth 1000 words for rendering bugs. Drop it in
-this comment box — GitHub uploads attachments inline.

.github/ISSUE_TEMPLATE/feature_request.md

@@ -1,20 +0,0 @@
----
-name: Feature request
-about: Propose a new behavior or enhancement
-labels: enhancement
----
-
-**Problem**
-What real problem does this solve? Who hits it?
-
-**Proposed change**
-What you'd like reasonix to do differently. Include a sketch of the
-UX or API if relevant.
-
-**Alternatives considered**
-What else you tried or thought about. "I just want it" is not an
-alternative.
-
-**Scope check**
-- [ ] This belongs in core reasonix (not better as a separate npm package)
-- [ ] I've read CLAUDE.md and CONTRIBUTING.md

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,29 @@
+name: Feature request
+description: Suggest an idea or improvement
+title: "[Feature]: "
+labels: ["enhancement"]
+body:
+  - type: dropdown
+    id: version-line
+    attributes:
+      label: Version line
+      description: Which line is this for?
+      options:
+        - "v2 — Go rewrite (1.x), main-v2 (active development)"
+        - "v1 — Legacy TypeScript (0.x), maintenance only"
+    validations:
+      required: true
+  - type: textarea
+    id: problem
+    attributes:
+      label: What problem does this solve?
+      description: What are you trying to do, and what gets in the way today?
+    validations:
+      required: true
+  - type: textarea
+    id: proposal
+    attributes:
+      label: Proposed solution
+      description: How would you like it to work?
+    validations:
+      required: false

.github/PULL_REQUEST_TEMPLATE.md

@@ -1,20 +0,0 @@
-<!-- Read CONTRIBUTING.md first if this is your first PR. -->
-
-## What
-
-<!-- One paragraph: what does this change? -->
-
-## Why
-
-<!-- Bug fix? Pre-existing issue? Linked discussion? -->
-
-## How to verify
-
-<!-- Steps the reviewer can run. `npm run verify` is assumed. -->
-
-## Checklist
-
-- [ ] `npm run verify` passes locally (lint + typecheck + tests + comment-policy gate)
-- [ ] No `Co-Authored-By: Claude` trailer in commits
-- [ ] Comments follow CONTRIBUTING.md (no module-essay headers, no incident history)
-- [ ] No edits to `CHANGELOG.md` — release notes are maintainer-written at release time

.github/issue-labeler.yml

@@ -1,25 +0,0 @@
-bug:
-  - '/\b(crash|crashed|crashes|panic|panicked|stack trace|stacktrace|traceback)\b/i'
-  - '/\b(broken|doesn''?t work|not working|stopped working|regression)\b/i'
-  - '/\b(throws? (an? )?error|uncaught|unhandled)\b/i'
-
-enhancement:
-  - '/\b(feature request|would (be nice|like|love)|please add|please support)\b/i'
-  - '/\b(add (a |the )?(flag|option|config|setting|switch)|allow (configuring|disabling|enabling))\b/i'
-  - '/\b(can (we|you|reasonix) (add|support)|proposal|propose adding)\b/i'
-
-documentation:
-  - '/\b(docs?\b|documentation|readme|getting started|tutorial|wiki|man page)\b/i'
-  - '/\b(missing (docs?|documentation)|undocumented)\b/i'
-
-question:
-  - '/\b(how (do|to|can) (i|we|you)|is it possible|is there a way|why does|why is)\b/i'
-  - '/\b(question:|q:)\s/i'
-
-dashboard:
-  - '/\b(dashboard|\/dashboard|web UI|web companion|browser)\b/i'
-  - '/\blocalhost:|127\.0\.0\.1/i'
-
-rendering:
-  - '/\b(flicker|flickering|render artifact|redraw|repaint|ghost row|ghosting|screen tear)\b/i'
-  - '/\b(TUI (glitch|artifact|garbled)|terminal (glitch|garbled|broken display))\b/i'

.github/workflows/ci.yml

@@ -2,63 +2,180 @@ name: CI
 
 on:
   push:
-    branches: [main]
+    branches: [main-v2]
   pull_request:
-    branches: [main]
+    branches: [main-v2]
+
+permissions:
+  contents: read
+
+concurrency:
+  group: ci-${{ github.ref }}
+  cancel-in-progress: true
 
 jobs:
-  build:
-    name: build (${{ matrix.os }} · node ${{ matrix.node }})
-    runs-on: ${{ matrix.os }}
+  test:
     strategy:
       fail-fast: false
       matrix:
-        os: [ubuntu-latest, windows-latest]
-        node: ["22"]
+        os: [ubuntu-latest, macos-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v4
 
-      - name: Setup Node ${{ matrix.node }}
-        uses: actions/setup-node@v4
+      - uses: actions/setup-go@v5
         with:
-          node-version: ${{ matrix.node }}
-          cache: npm
+          go-version-file: go.mod
+          cache: true
 
-      - name: Install dependencies
-        run: npm ci
-
-      - name: Lint (biome)
-        run: npm run lint
-
-      - name: Typecheck
-        run: npm run typecheck
-
-      - name: Build (tsup + dashboard)
-        run: npm run build
-
-      - name: Test (vitest + coverage)
-        run: npm run test:coverage
-
-      - name: Coverage job summary
-        if: always()
-        run: node scripts/coverage-summary.mjs
-
-      # Smoke-test the bench harnesses themselves. --dry skips all LLM
-      # calls, so this catches wiring regressions (task factories, CLI
-      # parsing, file IO, checker determinism) without needing a
-      # DEEPSEEK_API_KEY in CI. RUNNER_TEMP is set on every GH Actions
-      # runner (POSIX + Windows) — beats hardcoding /tmp.
-      - name: τ-bench harness dry-run
-        run: npx tsx benchmarks/tau-bench/runner.ts --dry --out "${{ runner.temp }}/tau-dry.json"
-
-      - name: Notify Discord on main failure
-        if: failure() && github.event_name == 'push' && github.ref == 'refs/heads/main'
-        shell: bash
-        env:
-          WEBHOOK_URL: ${{ secrets.DISCORD_CI_WEBHOOK }}
+      # Skipped on Windows: the runner checks out CRLF, so gofmt -l flags every
+      # file. gofmt output is OS-independent, so the Unix legs already cover it.
+      - name: gofmt
+        if: runner.os != 'Windows'
         run: |
-          test -n "$WEBHOOK_URL" || { echo "DISCORD_CI_WEBHOOK unset, skipping"; exit 0; }
-          RUN_URL="${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
-          curl -sS -X POST -H "Content-Type: application/json" \
-            -d "{\"content\":\"❌ CI failed on \`${{ github.ref_name }}\` (${{ matrix.os }}) — ${{ github.actor }}\n${RUN_URL}\"}" \
-            "$WEBHOOK_URL"
+          # Root module only — desktop/ is a separate module with its own tooling.
+          unformatted=$(gofmt -l . | grep -v '^desktop/' || true)
+          if [ -n "$unformatted" ]; then
+            echo "These files are not gofmt-clean:"
+            echo "$unformatted"
+            exit 1
+          fi
+
+      - name: vet
+        run: go vet ./...
+
+      - name: build
+        run: go build ./...
+
+      - name: test
+        env:
+          # Run the prompt-cache prefix-stability guard (TestCacheHit*) in CI: a
+          # regression there silently tanks the cache hit rate the project is
+          # built around.
+          REASONIX_RELEASE_CACHE_GUARD: "1"
+        run: go test ./...
+
+  race:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      # The matrix never runs -race (it needs cgo); the project's concurrency
+      # (plugin fan-out, background phase B, jobs Kill/Wait) would otherwise
+      # ship without race coverage.
+      - name: test -race
+        env:
+          REASONIX_RELEASE_CACHE_GUARD: "1"
+        run: go test -race ./...
+
+  desktop:
+    runs-on: ubuntu-22.04
+    defaults:
+      run:
+        working-directory: desktop
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: desktop/go.mod
+          cache: true
+          cache-dependency-path: desktop/go.sum
+
+      # The Go main package embeds frontend/dist; a placeholder lets the Go
+      # build/test resolve the embed without a full frontend build.
+      - name: Stub embedded frontend
+        run: |
+          mkdir -p frontend/dist
+          echo '<!doctype html><title>ci</title>' > frontend/dist/index.html
+
+      - name: gofmt
+        run: |
+          unformatted=$(gofmt -l .)
+          if [ -n "$unformatted" ]; then
+            echo "These files are not gofmt-clean:"
+            echo "$unformatted"
+            exit 1
+          fi
+
+      - name: go.mod tidy
+        run: |
+          go mod tidy
+          if ! git diff --quiet -- go.mod go.sum; then
+            echo "desktop/go.mod or go.sum is stale - run 'cd desktop && go mod tidy' and commit."
+            git diff -- go.mod go.sum
+            exit 1
+          fi
+
+      # WebKitGTK 4.0 toolchain (pinned to ubuntu-22.04; no webkit2_41 tag).
+      - name: Install Linux build deps
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y gcc libgtk-3-dev libwebkit2gtk-4.0-dev
+
+      - name: vet
+        run: go vet ./...
+
+      - name: build
+        run: go build ./...
+
+      - name: test
+        run: go test ./...
+
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      - name: golangci-lint
+        uses: golangci/golangci-lint-action@v7
+        with:
+          version: v2.12.2
+          args: --timeout=5m
+
+  govulncheck:
+    runs-on: ubuntu-latest
+    continue-on-error: true # informational — stdlib vulns need a Go patch release
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      - name: install govulncheck
+        run: go install golang.org/x/vuln/cmd/govulncheck@latest
+
+      - name: govulncheck
+        run: govulncheck ./...
+
+  coverage:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      - name: test with coverage
+        run: go test -coverprofile=coverage.out -covermode=atomic ./...
+
+      - name: upload coverage
+        uses: actions/upload-artifact@v4
+        with:
+          name: coverage-report
+          path: coverage.out
+          retention-days: 7

.github/workflows/codeql.yml

@@ -1,38 +0,0 @@
-name: CodeQL
-
-on:
-  push:
-    branches: [main]
-  pull_request:
-    branches: [main]
-  schedule:
-    - cron: "0 6 * * 1"
-
-jobs:
-  analyze:
-    name: analyze (${{ matrix.language }})
-    runs-on: ubuntu-latest
-    permissions:
-      actions: read
-      contents: read
-      security-events: write
-    strategy:
-      fail-fast: false
-      matrix:
-        language: [javascript-typescript]
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: Initialize CodeQL
-        uses: github/codeql-action/init@v3
-        with:
-          languages: ${{ matrix.language }}
-          queries: security-extended
-
-      - name: Autobuild
-        uses: github/codeql-action/autobuild@v3
-
-      - name: Analyze
-        uses: github/codeql-action/analyze@v3
-        with:
-          category: "/language:${{ matrix.language }}"

.github/workflows/e2e-bot.yml

@@ -0,0 +1,149 @@
+name: e2e-bot
+
+# Comment "/e2e" (fixed suite) or "/e2e diff" (generate tests for the PR's diff)
+# on a pull request to run the e2e benchmark against the real provider and post a
+# report back. Gated to trusted authors: the job checks out PR-head code and runs
+# it with the provider API key, so only the repo owner, members, and collaborators
+# may trigger it.
+
+on:
+  issue_comment:
+    types: [created]
+
+permissions:
+  contents: read
+  pull-requests: write
+
+concurrency:
+  group: e2e-bot-${{ github.event.issue.number }}
+  cancel-in-progress: true
+
+jobs:
+  e2e:
+    if: >-
+      github.event.issue.pull_request &&
+      contains(github.event.comment.body, '/e2e') &&
+      contains(fromJSON('["OWNER","MEMBER","COLLABORATOR"]'), github.event.comment.author_association)
+    runs-on: ubuntu-latest
+    steps:
+      - name: Acknowledge
+        uses: actions/github-script@v7
+        with:
+          script: |
+            await github.rest.reactions.createForIssueComment({
+              owner: context.repo.owner, repo: context.repo.repo,
+              comment_id: context.payload.comment.id, content: 'eyes',
+            });
+
+      # Default-branch checkout: this is where the harness (cmd/e2ebench), the
+      # suite, and a run --metrics-capable agent live.
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version: '1.22'
+          cache: true
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+
+      - name: Build harness + fallback agent from the default branch
+        # Harness (e2ebench) and suite always come from main-v2 so a PR can't weaken
+        # its own grader or tests. The agent is rebuilt from the PR head below; this
+        # main-v2 build is only the fallback for heads that predate `run --metrics`.
+        run: |
+          go build -o "$RUNNER_TEMP/reasonix-base" ./cmd/reasonix
+          go build -o "$RUNNER_TEMP/e2ebench" ./cmd/e2ebench
+          cp -r benchmarks/e2e "$RUNNER_TEMP/suite"
+
+      - name: Check out the PR head
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: gh pr checkout ${{ github.event.issue.number }}
+
+      - name: Build the agent from the PR head
+        # The whole point of the bot is to drive the PR's code, not main-v2's. Fall
+        # back to the main-v2 build only when the PR head can't yield a
+        # run --metrics-capable binary (build break or predates the flag).
+        id: agent
+        run: |
+          bin="$RUNNER_TEMP/reasonix-base"
+          src="main-v2 fallback (PR head lacks run --metrics)"
+          if go build -o "$RUNNER_TEMP/reasonix-pr" ./cmd/reasonix \
+            && "$RUNNER_TEMP/reasonix-pr" run -h 2>&1 | grep -q -- '-metrics'; then
+            bin="$RUNNER_TEMP/reasonix-pr"
+            src="PR head ($(git rev-parse --short HEAD))"
+          fi
+          echo "bin=$bin" >> "$GITHUB_OUTPUT"
+          echo "src=$src" >> "$GITHUB_OUTPUT"
+          echo "agent under test: $src"
+
+      - name: Write provider config
+        # User config covers suite tasks (they run in temp dirs); the repo-root copy
+        # covers diff mode (the agent runs in the repo root, where project config wins).
+        run: |
+          mkdir -p ~/.config/reasonix
+          cat > /tmp/reasonix-e2e.toml <<EOF
+          default_model = "e2e"
+
+          [[providers]]
+          name = "e2e"
+          kind = "openai"
+          base_url = "${{ vars.REASONIX_E2E_BASE_URL || 'https://api.deepseek.com' }}"
+          model = "${{ vars.REASONIX_E2E_MODEL || 'deepseek-v4-flash' }}"
+          api_key_env = "DEEPSEEK_API_KEY"
+          context_window = 20000   # small so the suite actually exercises compaction + cache churn
+
+          [providers.price]
+          cache_hit = ${{ vars.REASONIX_E2E_PRICE_CACHE_HIT || '0.02' }}
+          input = ${{ vars.REASONIX_E2E_PRICE_INPUT || '1' }}
+          output = ${{ vars.REASONIX_E2E_PRICE_OUTPUT || '2' }}
+          currency = "${{ vars.REASONIX_E2E_PRICE_CURRENCY || '¥' }}"
+
+          [permissions]
+          mode = "allow"
+
+          [codegraph]
+          enabled = false
+          EOF
+          cp /tmp/reasonix-e2e.toml ~/.config/reasonix/config.toml
+          cp /tmp/reasonix-e2e.toml ./reasonix.toml
+
+      - name: Run e2e
+        env:
+          DEEPSEEK_API_KEY: ${{ secrets.DEEPSEEK_API_KEY }}
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          COMMENT_BODY: ${{ github.event.comment.body }}
+        run: |
+          if [ -z "$DEEPSEEK_API_KEY" ]; then
+            printf 'missing DEEPSEEK_API_KEY secret\n\nAdd a repository secret named DEEPSEEK_API_KEY to enable the e2e bot.\n' > report.md
+            exit 0
+          fi
+          if printf '%s' "$COMMENT_BODY" | grep -q '/e2e[[:space:]]\+diff'; then
+            ATTEMPTS=$(printf '%s' "$COMMENT_BODY" | sed -nE 's@.*/e2e[[:space:]]+diff[[:space:]]+x([0-9]+).*@\1@p' | head -1)
+            [ -z "$ATTEMPTS" ] && ATTEMPTS=1
+            [ "$ATTEMPTS" -gt 5 ] && ATTEMPTS=5
+            BASE_REF=$(gh pr view ${{ github.event.issue.number }} --json baseRefName -q .baseRefName)
+            git fetch -q origin "$BASE_REF"
+            BASE=$(git merge-base "origin/$BASE_REF" HEAD)
+            "$RUNNER_TEMP/e2ebench" -mode diff -bin "${{ steps.agent.outputs.bin }}" -repo . -base "$BASE" -model e2e -attempts "$ATTEMPTS" -out report.md
+          else
+            "$RUNNER_TEMP/e2ebench" -bin "${{ steps.agent.outputs.bin }}" -suite "$RUNNER_TEMP/suite" -model e2e -out report.md -json report.json -budget 400000
+          fi
+
+      - name: Post report
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          TRIGGER_USER: ${{ github.event.comment.user.login }}
+        run: |
+          printf '\n> agent: %s · triggered by @%s\n' "${{ steps.agent.outputs.src }}" "$TRIGGER_USER" >> report.md
+          gh pr comment ${{ github.event.issue.number }} --body-file report.md
+
+      - name: Report failure
+        if: failure()
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: gh pr comment ${{ github.event.issue.number }} --body "🤖 e2e bot failed — see the [run log](${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }})."

.github/workflows/issue-labeler.yml

@@ -1,22 +0,0 @@
-name: Issue auto-label
-
-on:
-  issues:
-    types: [opened, edited]
-
-permissions:
-  issues: write
-  contents: read
-
-jobs:
-  label:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: github/[email protected]
-        with:
-          repo-token: ${{ secrets.GITHUB_TOKEN }}
-          configuration-path: .github/issue-labeler.yml
-          enable-versioned-regex: 0
-          include-title: 1
-          include-body: 1
-          sync-labels: 0

.github/workflows/issue-similarity.yml

@@ -1,20 +0,0 @@
-name: Issue similarity check
-
-on:
-  issues:
-    types: [opened, edited]
-
-permissions:
-  issues: write
-
-jobs:
-  similarity:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions-cool/issues-similarity-analysis@v1
-        with:
-          filter-threshold: 0.7
-          title-excludes: ""
-          comment-title: "### Possibly related issues"
-          comment-body: "- #${number} · ${title}"
-          show-footer: false

.github/workflows/issue-version-label.yml

@@ -0,0 +1,48 @@
+name: Label issue version
+
+# Issue forms can't let a reporter set a label directly (that needs triage rights),
+# so the bug/feature forms carry a "Version line" dropdown and this workflow reads
+# the submitted value and applies the matching v1/v2 label on their behalf.
+on:
+  issues:
+    types: [opened, edited]
+
+permissions:
+  issues: write
+
+concurrency:
+  group: issue-version-${{ github.event.issue.number }}
+  cancel-in-progress: true
+
+jobs:
+  label:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/github-script@v7
+        with:
+          script: |
+            const body = context.payload.issue.body || '';
+            // Pull the "Version line" section out of the rendered issue form.
+            const section = body.split(/^###\s+/m).find(s => /^Version line/i.test(s)) || '';
+            const m = section.match(/\bv([12])\b/i);
+            if (!m) {
+              core.info('No version line found; nothing to label.');
+              return;
+            }
+            const choice = 'v' + m[1];
+            const other = choice === 'v2' ? 'v1' : 'v2';
+            await github.rest.issues.addLabels({
+              ...context.repo,
+              issue_number: context.issue.number,
+              labels: [choice],
+            });
+            // Keep v1/v2 mutually exclusive in case an edit flipped the choice.
+            try {
+              await github.rest.issues.removeLabel({
+                ...context.repo,
+                issue_number: context.issue.number,
+                name: other,
+              });
+            } catch (e) {
+              core.info(`No ${other} label to remove.`);
+            }

.github/workflows/oosmetrics.yml

@@ -1,17 +0,0 @@
-name: oosmetrics Health Check
-
-on:
-  schedule:
-    - cron: "0 9 * * 1"
-  workflow_dispatch:
-
-permissions:
-  issues: write
-
-jobs:
-  report:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: oosmetrics/health-check-action@v1
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/pages.yml

@@ -0,0 +1,45 @@
+name: Deploy site
+
+on:
+  push:
+    branches: [main-v2]
+    paths: ['site/**', '.github/workflows/pages.yml']
+  workflow_dispatch:
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+concurrency:
+  group: pages
+  cancel-in-progress: false
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-node@v4
+        with:
+          node-version: '22'
+          cache: npm
+          cache-dependency-path: site/package-lock.json
+      - name: Build
+        working-directory: site
+        run: |
+          npm ci
+          npm run build
+      - uses: actions/upload-pages-artifact@v3
+        with:
+          path: site/dist
+
+  deploy:
+    needs: build
+    runs-on: ubuntu-latest
+    environment:
+      name: github-pages
+      url: ${{ steps.deploy.outputs.page_url }}
+    steps:
+      - id: deploy
+        uses: actions/deploy-pages@v4

.github/workflows/pr-version-label.yml

@@ -0,0 +1,49 @@
+name: Label PR version
+
+# Contributors (especially from forks) can't set labels themselves — that needs
+# triage rights. A PR's version line is fully determined by its base branch, so
+# this reads base.ref and applies v1/v2 on their behalf. pull_request_target runs
+# in the base repo's context so it has write access even for fork PRs; it only
+# reads trusted metadata (the base ref) and never checks out or runs PR code.
+on:
+  pull_request_target:
+    types: [opened, reopened, edited]
+
+permissions:
+  pull-requests: write
+
+concurrency:
+  group: pr-version-${{ github.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  label:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/github-script@v7
+        with:
+          script: |
+            const base = context.payload.pull_request.base.ref;
+            const label = base === 'v1' ? 'v1'
+                        : base === 'main-v2' ? 'v2'
+                        : null;
+            if (!label) {
+              core.info(`Base branch '${base}' isn't a release line; skipping.`);
+              return;
+            }
+            const other = label === 'v2' ? 'v1' : 'v2';
+            await github.rest.issues.addLabels({
+              ...context.repo,
+              issue_number: context.payload.pull_request.number,
+              labels: [label],
+            });
+            // Drop the other line's label if the base was changed after opening.
+            try {
+              await github.rest.issues.removeLabel({
+                ...context.repo,
+                issue_number: context.payload.pull_request.number,
+                name: other,
+              });
+            } catch (e) {
+              core.info(`No ${other} label to remove.`);
+            }

.github/workflows/publish-dsnix.yml

@@ -1,63 +0,0 @@
-name: Publish dsnix
-
-# Tag namespace: `dsnix-v<semver>` triggers npm publish for the
-# dsnix shim package only. Plain `v<semver>` tags publish the main
-# `reasonix` CLI (see publish-npm.yml); the two are decoupled so the
-# alias can be released independently when the shim itself changes.
-on:
-  push:
-    tags: ["dsnix-v*"]
-  workflow_dispatch:
-    inputs:
-      tag:
-        description: "Tag to publish (e.g. dsnix-v0.48.0). Required when running manually."
-        required: true
-        type: string
-
-permissions:
-  contents: read
-  id-token: write
-
-jobs:
-  publish:
-    name: publish dsnix to npm
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          ref: ${{ github.event.inputs.tag || github.ref }}
-
-      - name: Setup Node 22
-        uses: actions/setup-node@v4
-        with:
-          node-version: "22"
-          cache: npm
-          registry-url: https://registry.npmjs.org/
-
-      - name: Verify tag matches packages/dsnix/package.json version
-        shell: bash
-        run: |
-          tag="${{ github.event.inputs.tag || github.ref_name }}"
-          tag_version="${tag#dsnix-v}"
-          pkg_version=$(node -p "require('./packages/dsnix/package.json').version")
-          if [ "$tag_version" != "$pkg_version" ]; then
-            echo "::error::tag $tag (version $tag_version) does not match packages/dsnix/package.json version $pkg_version"
-            exit 1
-          fi
-          echo "tag $tag matches packages/dsnix/package.json $pkg_version"
-
-      - name: Verify reasonix@<version> is already published on npm
-        shell: bash
-        run: |
-          dep_version=$(node -p "require('./packages/dsnix/package.json').dependencies.reasonix")
-          if ! npm view "reasonix@${dep_version}" version >/dev/null 2>&1; then
-            echo "::error::dsnix depends on reasonix@${dep_version} but that version is not published yet"
-            exit 1
-          fi
-          echo "reasonix@${dep_version} is published — safe to publish dsnix"
-
-      - name: Publish dsnix to npm
-        run: npm publish --access public
-        working-directory: packages/dsnix
-        env:
-          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}

.github/workflows/publish-npm.yml

@@ -1,66 +0,0 @@
-name: Publish npm
-
-# Tag namespace: plain `v<semver>` triggers npm publish for the
-# reasonix CLI package. Desktop Tauri releases use `desktop-v*`
-# (see release.yml) — pushing one of those never reaches this
-# workflow, so the two release tracks are fully decoupled.
-on:
-  push:
-    tags: ["v*"]
-  workflow_dispatch:
-    inputs:
-      tag:
-        description: "Tag to publish (e.g. v0.47.0). Required when running manually."
-        required: true
-        type: string
-
-permissions:
-  contents: read
-  # `id-token: write` is the npm provenance handshake — harmless if we
-  # don't enable provenance, required if we ever do.
-  id-token: write
-
-jobs:
-  publish:
-    name: publish to npm
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          # workflow_dispatch lets a maintainer publish an older tag
-          # without re-pushing it; tag pushes already point HEAD at the
-          # right ref so the fallback is a no-op.
-          ref: ${{ github.event.inputs.tag || github.ref }}
-
-      - name: Setup Node 22
-        uses: actions/setup-node@v4
-        with:
-          node-version: "22"
-          cache: npm
-          registry-url: https://registry.npmjs.org/
-
-      # Fail-fast if someone tags without bumping package.json. npm itself
-      # would reject the duplicate publish, but the error there is
-      # opaque ("403 You cannot publish over the previously published
-      # versions") and CI is the place to catch it cleanly.
-      - name: Verify tag matches package.json version
-        shell: bash
-        run: |
-          tag="${{ github.event.inputs.tag || github.ref_name }}"
-          tag_version="${tag#v}"
-          pkg_version=$(node -p "require('./package.json').version")
-          if [ "$tag_version" != "$pkg_version" ]; then
-            echo "::error::tag $tag (version $tag_version) does not match package.json version $pkg_version"
-            exit 1
-          fi
-          echo "tag $tag matches package.json $pkg_version"
-
-      - name: Install dependencies
-        run: npm ci
-
-      # `npm publish` runs the prepublishOnly hook (lint + typecheck +
-      # test + build) automatically, so no separate verify step here.
-      - name: Publish to npm
-        run: npm publish --access public
-        env:
-          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}

.github/workflows/release-desktop.yml

@@ -0,0 +1,225 @@
+name: Release desktop
+
+# Desktop (Wails) release line. Tag namespace `desktop-v<semver>` triggers this —
+# plain `v<semver>` tags are the CLI release (release.yml) and intentionally do NOT
+# trigger here (GitHub glob `v*` does not match `desktop-v*`). Bump with:
+#   git tag desktop-vX.Y.Z && git push origin desktop-vX.Y.Z
+#
+# Wails cannot cross-compile a CGO/WebKit binary, so build/ fans out to one native
+# runner per platform. Artifacts are minisign-signed (MINISIGN_* secrets), a
+# latest.json manifest is generated, and everything is published to a GitHub
+# release and mirrored to R2 (the updater reads R2 first, GitHub as fallback).
+on:
+  push:
+    tags: ["desktop-v*"]
+  workflow_dispatch:
+    inputs:
+      tag:
+        description: "Tag to publish (e.g. desktop-v1.1.0). Required when run manually."
+        required: true
+        type: string
+
+permissions:
+  contents: write # create the release and upload artifacts
+
+jobs:
+  cache-guard:
+    name: cache hit guard
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      - name: Cache hit guard
+        run: ./scripts/cache-guard.sh
+
+  build:
+    name: build (${{ matrix.name }})
+    needs: cache-guard
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          # One universal macOS build (Intel + Apple Silicon) on a single arm64
+          # runner — avoids the scarce/slow macos-13 Intel runner.
+          - { runner: macos-14, platform: darwin/universal, name: darwin-universal }
+          - { runner: windows-latest, platform: windows/amd64, name: windows-amd64 }
+          - { runner: ubuntu-22.04, platform: linux/amd64, name: linux-amd64 }
+    runs-on: ${{ matrix.runner }}
+    defaults:
+      run:
+        shell: bash # desktop-build.sh is bash; windows runners default to pwsh otherwise
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: desktop/go.mod
+          cache: true
+          cache-dependency-path: desktop/go.sum
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+      - uses: pnpm/action-setup@v4
+        with:
+          version: 10
+
+      # Linux: WebKitGTK 4.1 toolchain (-tags webkit2_41 in desktop-build.sh).
+      # 4.1 ships from ubuntu-22.04 on and is the only one present on 24.04+/Fedora 40+.
+      - name: Install Linux build deps
+        if: runner.os == 'Linux'
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y gcc libgtk-3-dev libwebkit2gtk-4.1-dev
+
+      # Windows: NSIS provides makensis for `wails build -nsis`.
+      - name: Install NSIS
+        if: runner.os == 'Windows'
+        run: |
+          choco install nsis -y --no-progress
+          echo "C:\Program Files (x86)\NSIS" >> "$GITHUB_PATH"
+
+      # macOS: create-dmg packages the .app into a drag-to-Applications .dmg.
+      - name: Install create-dmg
+        if: runner.os == 'macOS'
+        run: brew install create-dmg
+
+      - name: Install Wails CLI
+        run: go install github.com/wailsapp/wails/v2/cmd/[email protected]
+
+      - name: Resolve version
+        id: ver
+        run: |
+          TAG="${{ github.event_name == 'workflow_dispatch' && inputs.tag || github.ref_name }}"
+          echo "tag=$TAG" >> "$GITHUB_OUTPUT"
+          echo "version=${TAG#desktop-}" >> "$GITHUB_OUTPUT"
+
+      - name: Build and package
+        run: scripts/desktop-build.sh "${{ matrix.platform }}" "${{ steps.ver.outputs.version }}"
+
+      - name: Sign artifacts (minisign)
+        working-directory: desktop
+        env:
+          MINISIGN_PRIVATE_KEY: ${{ secrets.MINISIGN_PRIVATE_KEY }}
+          MINISIGN_PASSWORD: ${{ secrets.MINISIGN_PASSWORD }}
+        run: go run ./cmd/sign sign ../dist/*
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist-${{ matrix.name }}
+          path: dist/*
+          if-no-files-found: error
+
+  publish:
+    name: publish release
+    needs: build
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: desktop/go.mod
+          cache: true
+          cache-dependency-path: desktop/go.sum
+
+      - uses: actions/download-artifact@v4
+        with:
+          path: dist
+          pattern: dist-*
+          merge-multiple: true
+
+      - name: Resolve version
+        id: ver
+        run: |
+          TAG="${{ github.event_name == 'workflow_dispatch' && inputs.tag || github.ref_name }}"
+          VERSION="${TAG#desktop-}"
+          echo "tag=$TAG" >> "$GITHUB_OUTPUT"
+          echo "version=$VERSION" >> "$GITHUB_OUTPUT"
+          # A version with a suffix (v1.0.0-rc1) is a prerelease.
+          case "$VERSION" in *-*) echo "prerelease=true" >> "$GITHUB_OUTPUT";; *) echo "prerelease=false" >> "$GITHUB_OUTPUT";; esac
+
+      # Generate latest.json with GitHub release download URLs; the mirror step
+      # rewrites them to R2 afterwards. GITHUB_REPOSITORY is provided by the runner.
+      - name: Generate manifest
+        working-directory: desktop
+        run: go run ./cmd/sign manifest ../dist "${{ steps.ver.outputs.version }}" "${{ steps.ver.outputs.tag }}"
+
+      - name: Publish GitHub release
+        env:
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          args=(--title "Reasonix Desktop ${{ steps.ver.outputs.version }}" --generate-notes)
+          [ "${{ steps.ver.outputs.prerelease }}" = "true" ] && args+=(--prerelease)
+          gh release create "${{ steps.ver.outputs.tag }}" dist/* "${args[@]}"
+
+  mirror:
+    name: mirror to R2
+    needs: publish
+    runs-on: ubuntu-latest
+    # Skip cleanly when R2 isn't configured; GitHub release still works as fallback.
+    if: ${{ github.repository_owner == 'esengine' }}
+    env:
+      HAS_R2: ${{ secrets.R2_ACCESS_KEY_ID != '' && secrets.R2_SECRET_ACCESS_KEY != '' && secrets.R2_ACCOUNT_ID != '' && secrets.R2_BUCKET != '' }}
+    steps:
+      - name: Resolve tag
+        id: ver
+        run: |
+          TAG="${{ github.event_name == 'workflow_dispatch' && inputs.tag || github.ref_name }}"
+          echo "tag=$TAG" >> "$GITHUB_OUTPUT"
+          case "${TAG#desktop-}" in *-*) echo "prerelease=true" >> "$GITHUB_OUTPUT";; *) echo "prerelease=false" >> "$GITHUB_OUTPUT";; esac
+
+      - name: Download release assets
+        if: env.HAS_R2 == 'true'
+        env:
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          mkdir -p assets
+          gh release download "${{ steps.ver.outputs.tag }}" -R "${{ github.repository }}" -D assets
+
+      # Rewrite both url and sig inside latest.json from github.com to the R2 CDN,
+      # so the updater pulls the manifest AND the heavy artifacts from R2.
+      - name: Rewrite latest.json URLs to R2
+        if: env.HAS_R2 == 'true'
+        env:
+          R2_PUBLIC_BASE: https://pub-147fb53b9c1e4bbf891a257968619ea7.r2.dev
+          TAG: ${{ steps.ver.outputs.tag }}
+        run: |
+          f=assets/latest.json
+          jq --arg base "$R2_PUBLIC_BASE" --arg tag "$TAG" '
+            .platforms |= with_entries(
+              .value.url |= sub("https://github.com/[^/]+/[^/]+/releases/download/[^/]+/"; "\($base)/\($tag)/")
+              | .value.sig |= sub("https://github.com/[^/]+/[^/]+/releases/download/[^/]+/"; "\($base)/\($tag)/")
+            )
+          ' "$f" > "$f.new"
+          mv "$f.new" "$f"
+          cat "$f"
+
+      - name: Configure AWS CLI for R2
+        if: env.HAS_R2 == 'true'
+        run: |
+          aws configure set aws_access_key_id "${{ secrets.R2_ACCESS_KEY_ID }}"
+          aws configure set aws_secret_access_key "${{ secrets.R2_SECRET_ACCESS_KEY }}"
+          aws configure set region auto
+
+      - name: Mirror to R2
+        if: env.HAS_R2 == 'true'
+        env:
+          R2_ACCOUNT_ID: ${{ secrets.R2_ACCOUNT_ID }}
+          R2_BUCKET: ${{ secrets.R2_BUCKET }}
+          TAG: ${{ steps.ver.outputs.tag }}
+        run: |
+          ENDPOINT="https://${R2_ACCOUNT_ID}.r2.cloudflarestorage.com"
+          aws s3 cp assets/ "s3://${R2_BUCKET}/${TAG}/" --recursive --endpoint-url "$ENDPOINT"
+          # Prereleases publish to their own tag prefix but must not move the
+          # updater's latest/ pointer; stable releases do.
+          if [ "${{ steps.ver.outputs.prerelease }}" = "true" ]; then
+            echo "prerelease — leaving R2 latest/ untouched"
+          else
+            aws s3 cp assets/ "s3://${R2_BUCKET}/latest/" --recursive --endpoint-url "$ENDPOINT"
+          fi

.github/workflows/release-mirror.yml

@@ -1,103 +0,0 @@
-name: Release mirror
-
-on:
-  release:
-    types: [published]
-  workflow_dispatch:
-    inputs:
-      tag:
-        description: "Tag to mirror (e.g. v0.42.0)"
-        required: true
-        type: string
-
-permissions:
-  contents: read
-
-jobs:
-  mirror:
-    name: Mirror release assets to R2
-    runs-on: ubuntu-latest
-    env:
-      HAS_R2: ${{ secrets.R2_ACCESS_KEY_ID != '' && secrets.R2_SECRET_ACCESS_KEY != '' && secrets.R2_ACCOUNT_ID != '' && secrets.R2_BUCKET != '' }}
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: Resolve tag
-        id: tag
-        shell: bash
-        run: |
-          if [ "${{ github.event_name }}" = "workflow_dispatch" ]; then
-            echo "name=${{ inputs.tag }}" >> "$GITHUB_OUTPUT"
-          else
-            echo "name=${{ github.event.release.tag_name }}" >> "$GITHUB_OUTPUT"
-          fi
-
-      - name: Download release assets from GitHub
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        run: |
-          mkdir -p assets
-          gh release download "${{ steps.tag.outputs.name }}" \
-            -R "${{ github.repository }}" \
-            -D assets
-
-      # Rewrite bundle URLs inside latest.json (Tauri updater manifest)
-      # so the desktop pulls bundles from R2, not GitHub. Without this the
-      # updater fetches the small JSON from R2 but still downloads the
-      # heavy installer from github.com — defeating the mirror.
-      # No-op when latest.json is absent (CLI-only releases).
-      - name: Rewrite latest.json URLs → R2
-        if: env.HAS_R2 == 'true'
-        env:
-          R2_PUBLIC_BASE: https://pub-147fb53b9c1e4bbf891a257968619ea7.r2.dev
-          TAG: ${{ steps.tag.outputs.name }}
-        run: |
-          file=assets/latest.json
-          if [ ! -f "$file" ]; then
-            echo "no latest.json in release — skipping rewrite"
-            exit 0
-          fi
-          # GitHub URLs look like:
-          #   https://github.com/<owner>/<repo>/releases/download/<TAG>/<file>
-          # Rewrite to:
-          #   <R2_PUBLIC_BASE>/<TAG>/<file>
-          jq --arg base "$R2_PUBLIC_BASE" --arg tag "$TAG" '
-            .platforms |= with_entries(
-              .value.url |= sub("https://github.com/[^/]+/[^/]+/releases/download/[^/]+/"; "\($base)/\($tag)/")
-            )
-          ' "$file" > "$file.new"
-          mv "$file.new" "$file"
-          echo "rewrote latest.json:"
-          cat "$file"
-
-      - name: Configure AWS CLI for R2
-        if: env.HAS_R2 == 'true'
-        run: |
-          aws configure set aws_access_key_id "${{ secrets.R2_ACCESS_KEY_ID }}"
-          aws configure set aws_secret_access_key "${{ secrets.R2_SECRET_ACCESS_KEY }}"
-          aws configure set region auto
-
-      - name: Mirror to R2
-        if: env.HAS_R2 == 'true'
-        env:
-          R2_ACCOUNT_ID: ${{ secrets.R2_ACCOUNT_ID }}
-          R2_BUCKET: ${{ secrets.R2_BUCKET }}
-          TAG: ${{ steps.tag.outputs.name }}
-        run: |
-          ENDPOINT="https://${R2_ACCOUNT_ID}.r2.cloudflarestorage.com"
-          aws s3 cp assets/ "s3://${R2_BUCKET}/${TAG}/" \
-            --recursive \
-            --endpoint-url "$ENDPOINT"
-          aws s3 cp assets/ "s3://${R2_BUCKET}/latest/" \
-            --recursive \
-            --endpoint-url "$ENDPOINT"
-
-      - name: Summary
-        if: always()
-        run: |
-          {
-            echo "## Mirror result"
-            echo ""
-            echo "- Tag: \`${{ steps.tag.outputs.name }}\`"
-            echo "- R2: ${{ env.HAS_R2 == 'true' && '✓ mirrored' || 'skipped (no R2 secrets)' }}"
-          } >> "$GITHUB_STEP_SUMMARY"

.github/workflows/release.yml

@@ -1,187 +1,58 @@
-name: Release desktop
+name: Release
 
-# Tag namespace: `desktop-v<semver>` triggers desktop Tauri bundles.
-# Plain `v<semver>` tags are reserved for the npm CLI release (see
-# publish-npm.yml) — they intentionally do NOT trigger this workflow.
-# Bump the desktop with: `git tag desktop-vX.Y.Z && git push origin desktop-vX.Y.Z`.
 on:
   push:
-    tags: ["desktop-v*"]
-  workflow_dispatch:
-    inputs:
-      tag:
-        description: "Tag to publish (e.g. desktop-v0.47.0). Required when running manually."
-        required: true
-        type: string
+    tags: ['v*']
 
 permissions:
-  contents: write
+  contents: write # create the release and upload archives
 
 jobs:
-  bundle:
-    name: bundle (${{ matrix.target.label }})
-    strategy:
-      fail-fast: false
-      matrix:
-        target:
-          # macOS now ships two single-arch shards so Apple Silicon and Intel
-          # users get separate DMGs. `rust` and `tauri` stay aligned per shard.
-          - { os: ubuntu-22.04, label: "linux-x64", rust: "x86_64-unknown-linux-gnu", tauri: "x86_64-unknown-linux-gnu" }
-          - { os: macos-15-intel, label: "macos-x64", rust: "x86_64-apple-darwin", tauri: "x86_64-apple-darwin" }
-          - { os: macos-14, label: "macos-arm64", rust: "aarch64-apple-darwin", tauri: "aarch64-apple-darwin" }
-          - { os: windows-latest, label: "windows-x64", rust: "x86_64-pc-windows-msvc", tauri: "x86_64-pc-windows-msvc" }
-    runs-on: ${{ matrix.target.os }}
-    # secrets.* isn't accessible from step-level if:, so expose presence as a
-    # job-level env. Treat "true"/"false" strings as booleans downstream.
-    env:
-      HAS_APPLE_CERT: ${{ secrets.APPLE_CERTIFICATE != '' }}
-      HAS_WIN_CERT: ${{ secrets.WINDOWS_CERTIFICATE != '' }}
+  cache-guard:
+    name: cache hit guard
+    runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-
-      - name: Resolve tag
-        id: tag
-        shell: bash
-        run: |
-          if [ "${{ github.event_name }}" = "workflow_dispatch" ]; then
-            echo "name=${{ inputs.tag }}" >> "$GITHUB_OUTPUT"
-          else
-            echo "name=${GITHUB_REF_NAME}" >> "$GITHUB_OUTPUT"
-          fi
-
-      - name: Setup Node 22
-        uses: actions/setup-node@v4
+      - uses: actions/setup-go@v5
         with:
-          node-version: "22"
-          cache: npm
+          go-version-file: go.mod
+          cache: true
+      - run: ./scripts/cache-guard.sh
 
-      - name: Setup Rust (${{ matrix.target.rust }})
-        uses: dtolnay/rust-toolchain@stable
+  goreleaser:
+    name: archives + checksums + homebrew tap
+    needs: cache-guard
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
         with:
-          targets: ${{ matrix.target.rust }}
-
-      - name: Rust cache
-        uses: Swatinem/rust-cache@v2
+          fetch-depth: 0
+      - uses: actions/setup-go@v5
         with:
-          workspaces: desktop/src-tauri -> target
-          key: ${{ matrix.target.label }}
-
-      - name: Install Linux bundle deps
-        if: runner.os == 'Linux'
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y \
-            libwebkit2gtk-4.1-dev \
-            libgtk-3-dev \
-            libayatana-appindicator3-dev \
-            librsvg2-dev \
-            patchelf
-
-      - name: Install root deps
-        run: npm ci
-
-      - name: Sync desktop version from tag
-        run: node scripts/sync-desktop-version.mjs "${{ steps.tag.outputs.name }}"
-
-      - name: Build CLI bundle (dist/)
-        run: npm run build
-
-      - name: Install desktop deps
-        working-directory: desktop
-        run: npm ci
-
-      - name: Download bundled Node runtime
-        working-directory: desktop
-        run: npm run bundle:node
-
-      - name: Verify bundled Node matches host arch (macOS)
-        if: startsWith(matrix.target.os, 'macos-')
-        working-directory: desktop
-        run: |
-          expected_arch=$(uname -m)
-          archs=$(lipo -archs src-tauri/binaries/node)
-          echo "node archs: $archs"
-          echo "expected arch: $expected_arch"
-          if [ "$archs" != "$expected_arch" ]; then
-            echo "Expected $expected_arch Node, got: $archs" >&2
-            exit 1
-          fi
-          codesign -d --entitlements :- src-tauri/binaries/node 2>&1 | grep -q "com.apple.security.inherit"
-
-      - name: Build & verify desktop frontend
-        working-directory: desktop
-        shell: bash
-        run: |
-          npm run build
-          test -f dist/index.html
-          ls dist/assets/*.js >/dev/null
-
-      # tauri-action signs OS bundles whenever the corresponding env vars are
-      # defined — including when they're defined to the empty string, which is
-      # what `secrets.UNSET` evaluates to. Gate each signing block on its
-      # certificate being non-empty so unsigned prereleases don't trip
-      # `security import` / `signtool` on missing certs.
-      - name: Build Tauri bundle (unsigned)
-        if: >-
-          (matrix.target.os != 'windows-latest' || env.HAS_WIN_CERT != 'true')
-          && (!startsWith(matrix.target.os, 'macos-') || env.HAS_APPLE_CERT != 'true')
-        uses: tauri-apps/tauri-action@v0
+          go-version-file: go.mod
+          cache: true
+      - uses: goreleaser/goreleaser-action@v6
+        with:
+          version: '~> v2'
+          args: release --clean
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
-        with:
-          projectPath: desktop
-          tagName: ${{ steps.tag.outputs.name }}
-          releaseName: Reasonix ${{ steps.tag.outputs.name }}
-          releaseDraft: true
-          prerelease: false
-          args: --target ${{ matrix.target.tauri }}
+          HOMEBREW_TAP_TOKEN: ${{ secrets.HOMEBREW_TAP_TOKEN }}
 
-      - name: Build Tauri bundle (Windows, signed)
-        if: matrix.target.os == 'windows-latest' && env.HAS_WIN_CERT == 'true'
-        uses: tauri-apps/tauri-action@v0
+  npm:
+    name: publish npm packages
+    needs: cache-guard
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+      - uses: actions/setup-node@v4
+        with:
+          node-version: '22'
+          registry-url: 'https://registry.npmjs.org'
+      - run: node npm/build.mjs "${GITHUB_REF_NAME}" --publish
         env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
-          WINDOWS_CERTIFICATE: ${{ secrets.WINDOWS_CERTIFICATE }}
-          WINDOWS_CERTIFICATE_PASSWORD: ${{ secrets.WINDOWS_CERTIFICATE_PASSWORD }}
-        with:
-          projectPath: desktop
-          tagName: ${{ steps.tag.outputs.name }}
-          releaseName: Reasonix ${{ steps.tag.outputs.name }}
-          releaseDraft: true
-          prerelease: false
-          args: --target ${{ matrix.target.tauri }}
-
-      - name: Build Tauri bundle (macOS, signed + notarized)
-        if: startsWith(matrix.target.os, 'macos-') && env.HAS_APPLE_CERT == 'true'
-        uses: tauri-apps/tauri-action@v0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
-          APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
-          APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
-          APPLE_SIGNING_IDENTITY: ${{ secrets.APPLE_SIGNING_IDENTITY }}
-          APPLE_ID: ${{ secrets.APPLE_ID }}
-          APPLE_PASSWORD: ${{ secrets.APPLE_PASSWORD }}
-          APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
-        with:
-          projectPath: desktop
-          tagName: ${{ steps.tag.outputs.name }}
-          releaseName: Reasonix ${{ steps.tag.outputs.name }}
-          releaseDraft: true
-          prerelease: false
-          args: --target ${{ matrix.target.tauri }}
-
-      - name: Verify app Node inherits TCC grants (macOS)
-        if: startsWith(matrix.target.os, 'macos-')
-        working-directory: desktop
-        shell: bash
-        run: |
-          node_bin=$(find src-tauri/target -path '*/release/bundle/macos/*.app/Contents/Resources/node' -type f | head -n 1)
-          test -n "$node_bin"
-          codesign -vvv --strict "$node_bin"
-          codesign -d --entitlements :- "$node_bin" 2>&1 | grep -q "com.apple.security.inherit"
+          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}

.gitignore

@@ -1,55 +1,60 @@
-node_modules/
-dist/
-coverage/
-.stryker-tmp/
+# Build artifacts
+/bin/
+/dist/
+/stage/
+/npm/.stage/
+/reasonix
+/reasonix.exe
+
+# Website (Astro) build output
+/site/dist/
+/site/.astro/
+/site/_preview.png
+
+# Go test/build scratch
+*.test
+*.out
+*.coverprofile
+coverage.out
+cpu.prof
+mem.prof
+
+# Wails desktop artifacts
+/desktop/build/*
+!/desktop/build/appicon.png
+/desktop/desktop
+/desktop/frontend/dist/*
+!/desktop/frontend/dist/.gitkeep
+/desktop/frontend/wailsjs/
+/desktop/frontend/package.json.md5
+
+# Environment / local config (reasonix.example.toml is the shared template)
 .env
 .env.local
-*.log
-npm-debug.log*
-yarn-debug.log*
-yarn-error.log*
-.pnpm-debug.log*
+*.local
+/reasonix.toml
+/desktop/.env
+/desktop/reasonix.toml
+
+# Node / frontend dependencies and caches
+node_modules/
+.pnpm-store/
+.npm/
+
+# Project-local Reasonix state; keep the checked-in review command.
+/.reasonix/*
+!/.reasonix/commands/
+!/.reasonix/commands/review.md
+
+# CodeGraph per-project index (rebuilt locally; never committed)
+.codegraph/
+
+# Editor / OS
 .DS_Store
 Thumbs.db
 .idea/
 .vscode/
-*.tsbuildinfo
-transcripts/
-sessions/
-*.jsonl
-!tests/**/*.jsonl
-# Committed reference transcripts so readers can reasonix replay / diff
-# the v0.1 bench results without an API key.
-!benchmarks/tau-bench/transcripts/
-!benchmarks/tau-bench/transcripts/*.jsonl
-.turbo/
-# Local-only Claude Code settings — personal overrides, never committed.
-.claude/settings.local.json
-# Local agent notes — personal workflow reminders, never committed.
-agent.md
-# Per-user Reasonix state under .reasonix/ — committable team-level
-# files (settings.json, skills/) stay tracked, but the user-private
-# bits (semantic index, sessions, opt-out markers) never should.
-.reasonix/semantic/
-.reasonix/sessions/
-.reasonix/semantic-skip
-.reasonix/truncated-results/
-# Scratch entry regenerated each time scripts/bundle-codemirror.mjs runs.
-scripts/.cm-entry.mjs
-# Personal bun lockfile — project uses npm officially.
-bun.lock
 
-# Local portable Node/npm used for development on machines without npm.
-.tools/
-.npm-cache/
-
-# Tauri desktop shell build artifacts.
-desktop/src-tauri/target/
-desktop/src-tauri/gen/
-desktop/src-tauri/binaries/
-
-# Design mock-up reference (local, not part of repo)
-.design-ref/
-*.cpuprofile
-
-.probe-snapshots/
+# Scratch / agent working dirs (never committed)
+/tmp/
+/.codex/

.golangci.yml

@@ -0,0 +1,39 @@
+version: "2"
+
+linters:
+  enable:
+    - errcheck
+    - govet
+    - ineffassign
+    - staticcheck
+    - unused
+  settings:
+    errcheck:
+      check-type-assertions: false
+      exclude-functions:
+        - (*os.File).Close
+        - (io.Closer).Close
+        - (*net/http.Response).Body.Close
+        - golang.org/x/term.Restore
+    staticcheck:
+      checks:
+        - all
+        - -ST1005 # error strings ending with punctuation — too strict for initial adoption
+        - -ST1018 # Unicode format characters — intentional in TUI code
+        - -QF1001 # De Morgan's law — readability preference
+  exclusions:
+    generated: lax
+    presets:
+      - comments
+      - common-false-positives
+      - legacy
+      - std-error-handling
+    rules:
+      # Setup/teardown in tests routinely ignores cleanup errors.
+      - path: _test\.go
+        linters:
+          - errcheck
+
+issues:
+  max-issues-per-linter: 0
+  max-same-issues: 0

.goreleaser.yaml

@@ -0,0 +1,59 @@
+version: 2
+project_name: reasonix
+
+before:
+  hooks:
+    - go mod download
+
+builds:
+  - id: reasonix
+    main: ./cmd/reasonix
+    binary: reasonix
+    env:
+      - CGO_ENABLED=0
+    flags:
+      - -trimpath
+    ldflags:
+      - -s -w -X main.version={{ .Tag }}
+    goos: [darwin, linux, windows]
+    goarch: [amd64, arm64]
+
+archives:
+  - id: reasonix
+    ids: [reasonix]
+    name_template: "reasonix-{{ .Os }}-{{ .Arch }}"
+    formats: [tar.gz]
+    format_overrides:
+      - goos: windows
+        formats: [zip]
+
+checksum:
+  name_template: SHA256SUMS
+  algorithm: sha256
+
+homebrew_casks:
+  - name: reasonix
+    ids: [reasonix]
+    # Prereleases (v*-rc*) must not update the stable tap; brew has no separate
+    # prerelease channel, so a pushed rc cask becomes the default `brew install`.
+    skip_upload: "auto"
+    repository:
+      owner: esengine
+      name: homebrew-reasonix
+      branch: main
+      token: "{{ .Env.HOMEBREW_TAP_TOKEN }}"
+    homepage: "https://github.com/esengine/DeepSeek-Reasonix"
+    description: "Cache-first DeepSeek coding agent for the terminal."
+    commit_author:
+      name: reasonix
+      email: [email protected]
+    hooks:
+      post:
+        install: |
+          if OS.mac?
+            system_command "/usr/bin/xattr", args: ["-dr", "com.apple.quarantine", "#{staged_path}/reasonix"]
+          end
+
+release:
+  prerelease: auto
+  name_template: "Reasonix CLI {{ .Tag }}"

.reasonix/commands/review.md

@@ -0,0 +1,5 @@
+---
+description: Review a file for bugs
+argument-hint: [path]
+---
+Read $1 and list any correctness bugs or risky patterns, with file:line references, most important first. Focus: $ARGUMENTS.

CODE_OF_CONDUCT.md

@@ -1,83 +0,0 @@
-# Contributor Covenant Code of Conduct
-
-## Our Pledge
-
-We as members, contributors, and leaders pledge to make participation in our community a harassment-free experience for everyone, regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socio-economic status, nationality, personal appearance, race, caste, color, religion, or sexual identity and orientation.
-
-We pledge to act and interact in ways that contribute to an open, welcoming, diverse, inclusive, and healthy community.
-
-## Our Standards
-
-Examples of behavior that contributes to a positive environment for our community include:
-
-* Demonstrating empathy and kindness toward other people
-* Being respectful of differing opinions, viewpoints, and experiences
-* Giving and gracefully accepting constructive feedback
-* Accepting responsibility and apologizing to those affected by our mistakes, and learning from the experience
-* Focusing on what is best not just for us as individuals, but for the overall community
-
-Examples of unacceptable behavior include:
-
-* The use of sexualized language or imagery, and sexual attention or advances of any kind
-* Trolling, insulting or derogatory comments, and personal or political attacks
-* Public or private harassment
-* Publishing others' private information, such as a physical or email address, without their explicit permission
-* Other conduct which could reasonably be considered inappropriate in a professional setting
-
-## Enforcement Responsibilities
-
-Community leaders are responsible for clarifying and enforcing our standards of acceptable behavior and will take appropriate and fair corrective action in response to any behavior that they deem inappropriate, threatening, offensive, or harmful.
-
-Community leaders have the right and responsibility to remove, edit, or reject comments, commits, code, wiki edits, issues, and other contributions that are not aligned to this Code of Conduct, and will communicate reasons for moderation decisions when appropriate.
-
-## Scope
-
-This Code of Conduct applies within all community spaces, and also applies when an individual is officially representing the community in public spaces. Examples of representing our community include using an official e-mail address, posting via an official social media account, or acting as an appointed representative at an online or offline event.
-
-## Enforcement
-
-Instances of abusive, harassing, or otherwise unacceptable behavior may be reported to the project maintainer at <359807859@qq.com>. All complaints will be reviewed and investigated promptly and fairly.
-
-All community leaders are obligated to respect the privacy and security of the reporter of any incident.
-
-## Enforcement Guidelines
-
-Community leaders will follow these Community Impact Guidelines in determining the consequences for any action they deem in violation of this Code of Conduct:
-
-### 1. Correction
-
-**Community Impact**: Use of inappropriate language or other behavior deemed unprofessional or unwelcome in the community.
-
-**Consequence**: A private, written warning from community leaders, providing clarity around the nature of the violation and an explanation of why the behavior was inappropriate. A public apology may be requested.
-
-### 2. Warning
-
-**Community Impact**: A violation through a single incident or series of actions.
-
-**Consequence**: A warning with consequences for continued behavior. No interaction with the people involved, including unsolicited interaction with those enforcing the Code of Conduct, for a specified period of time. This includes avoiding interactions in community spaces as well as external channels like social media. Violating these terms may lead to a temporary or permanent ban.
-
-### 3. Temporary Ban
-
-**Community Impact**: A serious violation of community standards, including sustained inappropriate behavior.
-
-**Consequence**: A temporary ban from any sort of interaction or public communication with the community for a specified period of time. No public or private interaction with the people involved, including unsolicited interaction with those enforcing the Code of Conduct, is allowed during this period. Violating these terms may lead to a permanent ban.
-
-### 4. Permanent Ban
-
-**Community Impact**: Demonstrating a pattern of violation of community standards, including sustained inappropriate behavior, harassment of an individual, or aggression toward or disparagement of classes of individuals.
-
-**Consequence**: A permanent ban from any sort of public interaction within the community.
-
-## Attribution
-
-This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 2.1, available at [https://www.contributor-covenant.org/version/2/1/code_of_conduct.html][v2.1].
-
-Community Impact Guidelines were inspired by [Mozilla's code of conduct enforcement ladder][Mozilla CoC].
-
-For answers to common questions about this code of conduct, see the FAQ at [https://www.contributor-covenant.org/faq][FAQ]. Translations are available at [https://www.contributor-covenant.org/translations][translations].
-
-[homepage]: https://www.contributor-covenant.org
-[v2.1]: https://www.contributor-covenant.org/version/2/1/code_of_conduct.html
-[Mozilla CoC]: https://github.com/mozilla/diversity
-[FAQ]: https://www.contributor-covenant.org/faq
-[translations]: https://www.contributor-covenant.org/translations

CONTRIBUTING.md

@@ -1,152 +1,137 @@
 # Contributing to Reasonix
 
-Thanks for showing up. Reasonix is a small, opinionated codebase
-maintained primarily by [@esengine](https://github.com/esengine);
-PRs are welcome, but read this first so the round-trip is short.
+Thank you for your interest in contributing to Reasonix! This guide covers
+everything you need to get started.
 
-## Setup
+## Prerequisites
 
-```sh
-git clone https://github.com/esengine/reasonix
-cd reasonix
-npm install
-npm run dev          # tsx src/cli/index.ts — live source
+- **Go 1.25+** — the project targets the latest stable Go release
+- **Git** — for version control
+- **Node.js** (optional) — only if you work on the desktop app (`desktop/`)
+
+## Getting started
+
+```bash
+git clone https://github.com/esengine/DeepSeek-Reasonix.git
+cd DeepSeek-Reasonix
+go build ./cmd/reasonix    # builds the CLI binary
+go test ./...              # runs the full test suite
 ```
 
-Node ≥ 22. No global install needed during development.
+## Project structure
 
-For stack, layout, scripts, see [`REASONIX.md`](./REASONIX.md).
+| Directory | Purpose |
+|-----------|---------|
+| `cmd/reasonix` | CLI entry point |
+| `internal/agent` | Agent loop, session, coordinator |
+| `internal/cli` | TUI, subcommands, setup wizard |
+| `internal/control` | Transport-agnostic controller |
+| `internal/config` | TOML configuration loading |
+| `internal/tool/builtin` | Built-in tools (bash, read_file, …) |
+| `internal/provider` | Model-backend abstraction |
+| `internal/provider/openai` | OpenAI-compatible provider |
+| `internal/plugin` | MCP client (stdio + HTTP) |
+| `internal/event` | Typed event stream |
+| `internal/hook` | Shell hooks (PreToolUse, …) |
+| `internal/memory` | REASONIX.md hierarchy + auto-memory |
+| `internal/skill` | Skill discovery from Markdown |
+| `internal/sandbox` | OS-level sandboxing |
+| `internal/serve` | HTTP/SSE server frontend |
+| `internal/checkpoint` | Snapshot-based rewind |
+| `desktop/` | Wails-based desktop app (separate Go module) |
+| `docs/` | Engineering spec, migration guide |
 
-## Proposing changes
+### Dependency direction
 
-- **Bug fixes** — go ahead and open a PR. Include a reproduction.
-- **New features / behavior changes** — open an issue first to align
-  on scope and approach. Reasonix tries to stay small; "we could add
-  X" PRs that arrive cold are usually rejected or scoped down.
-- **External MCP servers, plugins, presets** — a thin wrapper is
-  fine; a sprawling integration is better hosted as a separate
-  package that depends on `reasonix`.
+```
+cli → {agent, plugin, config} → {tool, provider}
+```
 
-## Code rules
+Built-in subpackages import their parent to self-register via `init()`.
+Parents never import children.
 
-These are enforced by review and (where possible) by
-`tests/comment-policy.test.ts` — which runs under `npm run verify`
-and gates pre-push.
+## Development workflow
 
-### Comments — default is none
+### Building
 
-Write a comment ONLY when **why** is non-obvious and removing the
-comment would confuse a future reader. Justified examples:
+```bash
+make build          # go build ./...
+make test           # go test ./...
+make vet            # go vet ./...
+make fmt            # gofmt -w .
+make hooks          # install git hooks (pre-push: go vet)
+make cross          # cross-compile for all 6 targets
+```
 
-- a hidden constraint (`// Yoga miscounts wrap → must clamp to width-1`)
-- a workaround for a specific bug
-- a subtle invariant the type system can't express
+### Running tests
 
-Don't write:
+```bash
+go test ./...                           # all tests
+go test ./internal/agent/ -v            # verbose, one package
+go test ./internal/tool/builtin/ -run TestGrep  # one test
+```
 
-- **What the code does.** Names already say it. No `// when x is positive`
-  above `if (x > 0)`.
-- **Module-level essays.** Multi-paragraph docstrings at the top of a
-  file are dead weight. Two short lines max.
-- **Conversation history.** No "user reported X", "screenshot showed
-  Y", "v0.13.2 introduced Z". That belongs in commits / PR text.
-- **Section banners.** `// ─── helpers ───` is noise; group by export.
-- **Restated parameter docs.** If `function pad(f, top, right, bottom,
-  left)`, no `@param top - top padding`.
+### Code style
 
-If a comment is justified, **one line is almost always enough**.
-Comments needing 4+ lines usually mean the code itself needs to be
-clearer (rename, extract, simplify) before any comment is added.
+- `gofmt` is enforced by CI — format before committing
+- Follow existing patterns: wrap errors with `fmt.Errorf("...: %w", err)`
+- Library code never calls `os.Exit` or prints to stdout/stderr
+- Only `cli/` and `main/` decide exit codes and user-facing messages
+- Exported identifiers must have doc comments
 
-### TypeScript
+### Commit messages
 
-- Strict mode. No `any` without a `// biome-ignore` and a reason.
-- Prefer narrow types over option bags; if a function takes 5+
-  optional flags, split the responsibilities.
-- Don't re-export types just so two files can share them — move the
-  type to the file that owns the concept.
+Follow [Conventional Commits](https://www.conventionalcommits.org/):
 
-### Libraries over hand-rolled
+```
+feat(glob): add ** recursive pattern support
+fix: replace silent error discards with structured logging
+test(event): add comprehensive unit tests for event package
+docs: add CONTRIBUTING.md
+ci: add golangci-lint and govulncheck
+```
 
-If a problem has a well-maintained npm library, use it. Specific
-landmines this project has hit:
+## Adding a new built-in tool
 
-- Visual width / unicode width → `string-width`
-- Grapheme segmentation → `Intl.Segmenter`
-- ANSI strip → use what `string-width` ships with
-- Color → use `theme.ts` constants, not raw hex in component code
+1. Create `internal/tool/builtin/mytool.go`
+2. Implement the `tool.Tool` interface: `Name()`, `Description()`, `Schema()`, `ReadOnly()`, `Execute()`
+3. Register via `func init() { tool.RegisterBuiltin(myTool{}) }`
+4. Add tests in `internal/tool/builtin/builtin_test.go` or a separate `mytool_test.go`
+5. The tool is automatically available — `main` blank-imports `builtin`
 
-If a lib is missing a case, file the issue upstream and add a thin
-wrapper — don't fork a local table.
+## Adding a new model provider
 
-### Files
+(For MCP tool servers see `internal/plugin` instead — that's a different layer.)
 
-- One responsibility per file. New code goes in new files when an
-  existing one is already large.
-- File header comment: zero or one line.
-- No `index.ts` re-exports unless they meaningfully shrink the
-  public surface.
-- Don't create new `*.md` documentation files unless explicitly
-  asked.
+1. Create `internal/provider/myprovider/`
+2. Implement `provider.Provider`: `Name()`, `Stream()`
+3. Register via `func init() { provider.Register("mykind", New) }`
+4. The provider is available from config with `kind = "mykind"`
 
-### Errors / fallbacks
+## Adding i18n strings
 
-- Don't add try/catch for "internal" errors. Trust your own code.
-- Don't validate things the type system already proves.
-- Boundary code (user input, network, FS) does validate; everything
-  else trusts.
-- No "graceful fallback" silently masking bugs. Log + crash >
-  silent wrong output.
+1. Add the field to `internal/i18n/i18n.go` (`Messages` struct)
+2. Add the value in `internal/i18n/messages_en.go` and `messages_zh.go`
+3. The `TestCatalogsComplete` test will fail if you miss a locale
 
-### Tests
+## Submitting changes
 
-- Test what's hard to verify by reading the code: invariants, edge
-  cases, regressions.
-- Don't test type signatures or that `function returns X` (the type
-  system does that).
-- Don't write tests just to bump coverage.
+1. Fork the repository
+2. Create a feature branch from `main-v2`
+3. Make your changes with tests
+4. Ensure `go test ./...` passes
+5. Ensure `gofmt -l .` shows no changes
+6. Submit a pull request to `main-v2`
 
-### Git / commits
+## Reporting issues
 
-- Imperative mood, scope tag, why-not-what. See recent `git log`
-  for the pattern (`feat(ui): …`, `fix(loop): …`, `chore(release):
-  …`).
-- One logical change per commit; refactors land separately from
-  features.
-- No `Co-Authored-By: Claude` trailer.
+Open an issue on GitHub with:
+- Steps to reproduce
+- Expected vs actual behavior
+- Go version and OS
+- Relevant logs or error messages
 
-## PR expectations
+## License
 
-- Branch off `main`. One logical change per PR.
-- `npm run verify` must pass locally (lint + typecheck + tests +
-  comment-policy gate). Pre-push hook runs this; CI runs it on
-  Node 22.
-- Don't touch `CHANGELOG.md` — release notes are written by the
-  maintainer at release time, drawn from commit history. PR
-  descriptions are the authoritative record while the work is in
-  flight.
-
-## Code review
-
-Reasonix prefers blunt, fast review. Expect:
-
-- Line-level pushback on comments that explain *what* instead of *why*.
-- Pushback on new abstractions / flags introduced before there are
-  two real call sites.
-- Pushback on hand-rolled implementations of problems a maintained
-  npm library already solves.
-
-None of this is personal — it's how the codebase stays small.
-
-## Releasing (maintainers)
-
-1. Bump `package.json` version.
-2. Add `## [X.Y.Z] — <date>` to `CHANGELOG.md` with a hand-written
-   summary drawn from `git log` since the prior tag.
-3. `chore(release): X.Y.Z — <one-line summary>` commit.
-4. `git tag -a vX.Y.Z -m "..."`, push commit + tag.
-5. Wait for CI green, then `npm publish`.
-
-## Reporting security issues
-
-See [`SECURITY.md`](./SECURITY.md). Short version: don't open a public issue, email the maintainer privately.
+By contributing, you agree that your contributions will be licensed under the
+same license as the project.

Makefile

@@ -0,0 +1,52 @@
+VERSION := $(shell git describe --tags --always 2>/dev/null || echo dev)
+LDFLAGS := -s -w -X main.version=$(VERSION)
+GOEXE := $(shell go env GOEXE)
+
+# CodeGraph release pinned for the bundled MCP server / e2e test. Bump together
+# with any change to the integration in internal/codegraph.
+CODEGRAPH_VERSION := v0.9.7
+
+.PHONY: build vet fmt test hooks cross clean e2e-codegraph
+
+build:
+	CGO_ENABLED=0 go build -ldflags "$(LDFLAGS)" -o bin/reasonix$(GOEXE) ./cmd/reasonix
+	CGO_ENABLED=0 go build -ldflags "$(LDFLAGS)" -o bin/reasonix-plugin-example$(GOEXE) ./cmd/reasonix-plugin-example
+
+vet:
+	go vet ./...
+
+fmt:
+	gofmt -w .
+
+test:
+	go test ./...
+
+hooks:
+	@git config core.hooksPath .githooks
+	@echo "installed: core.hooksPath -> .githooks (pre-push runs go vet)"
+
+cross:
+	@mkdir -p dist
+	@for p in darwin/amd64 darwin/arm64 linux/amd64 linux/arm64 windows/amd64 windows/arm64; do \
+		os=$${p%/*}; arch=$${p#*/}; ext=; [ $$os = windows ] && ext=.exe; \
+		echo "build $$os/$$arch"; \
+		CGO_ENABLED=0 GOOS=$$os GOARCH=$$arch go build -ldflags "$(LDFLAGS)" -o dist/reasonix-$$os-$$arch$$ext ./cmd/reasonix; \
+	done
+
+clean:
+	rm -rf bin dist
+
+# Fetch the matching CodeGraph bundle into bin/codegraph/ (the distribution
+# layout: launcher at bin/codegraph/bin/codegraph beside bin/reasonix) and run the
+# gated MCP end-to-end test against it. Requires `gh`. Windows: install via the
+# upstream install.ps1 and run the test with REASONIX_CODEGRAPH_BIN set.
+e2e-codegraph:
+	@os=$$(uname -s | tr 'A-Z' 'a-z'); arch=$$(uname -m); \
+	case $$arch in arm64|aarch64) arch=arm64;; x86_64|amd64) arch=x64;; *) echo "unsupported arch $$arch"; exit 1;; esac; \
+	asset=codegraph-$$os-$$arch.tar.gz; dest=bin/codegraph; \
+	echo "fetching $$asset ($(CODEGRAPH_VERSION)) -> $$dest"; \
+	rm -rf $$dest && mkdir -p $$dest; \
+	gh release download $(CODEGRAPH_VERSION) -R colbymchenry/codegraph -p $$asset -O /tmp/$$asset; \
+	tar -xzf /tmp/$$asset -C $$dest --strip-components=1; \
+	REASONIX_CODEGRAPH_E2E=1 REASONIX_CODEGRAPH_BIN=$$PWD/$$dest/bin/codegraph \
+		go test ./internal/codegraph/ -run E2E -v -count=1

README.ja-JP.md

@@ -1,339 +0,0 @@
-<p align="center">
-  <img src="docs/logo.svg" alt="Reasonix" width="640"/>
-</p>
-
-> [!WARNING]
-> **これはレガシーの TypeScript 系列(Reasonix 0.x)で、メンテナンスのみです。**
-> 開発は **Go 版の書き直し** に移行しました —— 新しいデフォルトの [`main-v2`](https://github.com/esengine/DeepSeek-Reasonix/tree/main-v2) ブランチと[移行ガイド](https://github.com/esengine/DeepSeek-Reasonix/blob/main-v2/docs/MIGRATING.md)を参照してください。ここではバグ修正のみ反映されます。
-
-
-<p align="center">
-  <a href="./README.md">English</a>
-  &nbsp;·&nbsp;
-  <a href="./README.zh-CN.md">简体中文</a>
-  &nbsp;·&nbsp;
-  <strong>日本語</strong>
-  &nbsp;·&nbsp;
-  <a href="https://esengine.github.io/DeepSeek-Reasonix/">公式サイト</a>
-  &nbsp;·&nbsp;
-  <a href="https://esengine.github.io/DeepSeek-Reasonix/configuration.html">ガイド</a>
-  &nbsp;·&nbsp;
-  <a href="./docs/ARCHITECTURE.md">アーキテクチャ</a>
-  &nbsp;·&nbsp;
-  <a href="./benchmarks/">ベンチマーク</a>
-  &nbsp;·&nbsp;
-  <strong><a href="https://discord.gg/XF78rEME2D">Discord</a></strong>
-</p>
-
-<p align="center">
-  <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/v/reasonix.svg?style=flat-square&color=cb3837&labelColor=161b22&logo=npm&logoColor=white" alt="npm version"/></a>
-  <a href="https://github.com/esengine/reasonix/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/esengine/reasonix/ci.yml?style=flat-square&label=ci&labelColor=161b22&logo=githubactions&logoColor=white" alt="CI"/></a>
-  <a href="./LICENSE"><img src="https://img.shields.io/npm/l/reasonix.svg?style=flat-square&color=8b949e&labelColor=161b22" alt="license"/></a>
-  <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/dm/reasonix.svg?style=flat-square&color=3fb950&labelColor=161b22&label=downloads" alt="downloads"/></a>
-  <a href="./package.json"><img src="https://img.shields.io/node/v/reasonix.svg?style=flat-square&color=5fa04e&labelColor=161b22&logo=nodedotjs&logoColor=white" alt="node"/></a>
-  <a href="https://github.com/esengine/reasonix/stargazers"><img src="https://img.shields.io/github/stars/esengine/reasonix.svg?style=flat-square&color=dbab09&labelColor=161b22&logo=github&logoColor=white" alt="GitHub stars"/></a>
-  <a href="https://atomgit.com/esengine/DeepSeek-Reasonix"><img src="https://atomgit.com/esengine/DeepSeek-Reasonix/star/badge.svg" alt="AtomGit stars"/></a>
-  <a href="https://github.com/esengine/reasonix/graphs/contributors"><img src="https://img.shields.io/github/contributors/esengine/reasonix.svg?style=flat-square&color=bc8cff&labelColor=161b22&logo=github&logoColor=white" alt="contributors"/></a>
-  <a href="https://github.com/esengine/reasonix/discussions"><img src="https://img.shields.io/github/discussions/esengine/reasonix.svg?style=flat-square&color=58a6ff&labelColor=161b22&logo=github&logoColor=white" alt="Discussions"/></a>
-  <a href="https://discord.gg/XF78rEME2D"><img src="https://img.shields.io/badge/discord-join-5865F2.svg?style=flat-square&labelColor=161b22&logo=discord&logoColor=white" alt="Discord"/></a>
-</p>
-
-<p align="center">
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/9e931d80-2050-4b10-902e-44970cc133ad.svg" alt="oosmetrics — Top 2 in Agents by velocity"/></a>
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/556d94b3-61b7-486b-baf2-888b9327deab.svg" alt="oosmetrics — Top 3 in LLMs by velocity"/></a>
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/0f457d4c-efca-4d15-ad2b-139691ff342c.svg" alt="oosmetrics — Top 3 in CLI by velocity"/></a>
-</p>
-
-<br/>
-
-<h3 align="center">ターミナルで動く DeepSeek ネイティブの AI コーディングエージェント。</h3>
-<p align="center">プレフィックスキャッシュの安定性を中心に設計されているため、長時間のセッションでもトークンコストを低く保ち、動かしっぱなしにできます。</p>
-
-<br/>
-
-<p align="center">
-  <img src="docs/assets/hero-terminal.svg" alt="Reasonix code mode — assistant proposes a SEARCH/REPLACE edit; nothing on disk until /apply" width="860"/>
-</p>
-
-<br/>
-
-> [!TIP]
-> **キャッシュの安定性はオンにする「機能」ではなく、ループそのものが前提として設計された不変条件です。** これこそが Reasonix が DeepSeek 専用である理由であり、すべてのレイヤーがバイト単位で安定したプレフィックスキャッシュの仕組みに合わせて調整されています。
-
-> [!NOTE]
-> **実ユーザーの1日（2026-05-01）:** 入力トークン 4億3,500万、**キャッシュヒット率 99.82%**、同じワークロードを `v4-flash` でキャッシュなしに実行した場合の約 \$61 に対して、わずか約 \$12 で済んでいます — 詳しくは[ケーススタディ](./benchmarks/real-world-cache/README.md)を参照。キャッシュ可能なバイト列を提供するのは DeepSeek ですが、それを長時間のセッションにわたってキャッシュ可能に保つのが [Pillar 1](./docs/ARCHITECTURE.md#pillar-1--cache-first-loop) の4つの仕組みです。
-
-> [!IMPORTANT]
-> **Community · 加入社区** — セットアップのヘルプ（`#help` / `#求助`）、ワークフローのショーケース、機能アイデア、コントリビューター限定の PR 調整用チャンネルを備えたバイリンガル Discord。サーバー内で GitHub を認証すると、自動的に **Contributor** ロールが付与されます。→ **<https://discord.gg/XF78rEME2D>**
-
-<br/>
-
-## インストール
-
-Node ≥ 22 が必要です。macOS · Linux · Windows（PowerShell · Git Bash · Windows Terminal）で動作します。
-
-`reasonix` コマンドを `PATH` 上で使えるようにしたい場合は、Reasonix をグローバルにインストールします:
-
-~~~bash
-npm install -g reasonix
-reasonix code my-project   # 初回実行時に DeepSeek API キーを貼り付け。以降は保存される
-~~~
-
-または、グローバルインストールせずに一度だけ実行することもできます:
-
-~~~bash
-cd my-project
-npx reasonix code          # デフォルトでは常に最新のパッケージを使用
-~~~
-
-[DeepSeek API キーを取得 →](https://platform.deepseek.com/api_keys) · フラグは `reasonix code --help` を参照。
-
-Reasonix を毎日使うなら、グローバルインストールが最もシンプルな方法です。とりあえず試してみたいだけなら `npx` を使ってください。
-
-**キー入力を減らしたい？** より短い `dsnix` エイリアスが同じ CLI を指します:
-
-~~~bash
-npm install -g dsnix       # PATH 上に `dsnix` を公開。reasonix に依存
-npx dsnix@latest code      # より短いコマンドで一度だけ実行
-~~~
-
-`npm install -g reasonix` でグローバルインストールすると `dsnix` のシムも PATH 上に配置されるため、2つは相互に置き換え可能です。
-
-サブコマンドなしの `reasonix`（裸の `reasonix`）はカレントディレクトリで `code` を起動します — `reasonix` と入力するのは `reasonix code` と等価です。
-
-| コマンド | 用途 |
-|---|---|
-| `reasonix` / `reasonix code [dir]` | コーディングエージェント。**ここから始めてください。** |
-| `reasonix chat` | プレーンなチャット — ファイルシステムやシェルのツールなし。 |
-| `reasonix run "task"` | 一度だけ実行し、stdout にストリーミング。パイプ向き。 |
-| `reasonix doctor` | ヘルスチェック: Node、API キー、MCP の配線。 |
-| `reasonix update` | Reasonix 自体をアップグレード。 |
-
-その他のサブコマンド（`replay` · `diff` · `events` · `stats` · `index` · `mcp` · `prune-sessions`）は `reasonix --help` と [CLI リファレンス](https://esengine.github.io/DeepSeek-Reasonix/#cli)にあります。
-
-### QQ チャンネル
-
-QQ は既存の `chat`、`code`、またはデスクトップのセッションをリモートチャンネルとして拡張できます。これは独立したランタイムモードではなく、現在のセッションフローの一部です。
-
-- CLI: セッションを開始し、`/qq connect` を実行
-- デスクトップ: `Settings -> General -> QQ Channel` を開く
-
-接続すると、QQ のメッセージを現在のセッションに取り込み、アシスタントの返信を QQ に転送し、その後のやり取りをリモートで継続できます。
-
-詳しいセットアップ、デスクトップのクイックスタート、トラブルシューティングは [QQ チャンネルのセットアップ](./docs/qq-connect.md)を参照してください。
-
-### デスクトップクライアント（プレリリース）
-
-同じループの GUI が欲しいユーザー向けのネイティブ Tauri クライアント。マルチタブ対応で、右パネルにはこのセッションでエージェントが読み取り・編集したファイルが表示され、下部には同じコスト / キャッシュ / トークンのメーターがリアルタイムで表示されます。同じ DeepSeek API キー、同じ `~/.reasonix` 設定を使い、デスクトップ版は独自の Node ランタイムを同梱するため、別途 `npm install` する必要はありません。
-
-各プラットフォームのインストーラーは [GitHub Releases](https://github.com/esengine/DeepSeek-Reasonix/releases) からダウンロードできます。デスクトップ版は **プレリリース** として提供されます: ループとプロトコルは CLI と同じですが、UI はまだ磨き込み中で、インストーラーはまだコード署名されていません。
-
-- **macOS** — 初回起動時に Gatekeeper に引っかかります。一度きりの対処: `xattr -dr com.apple.quarantine /Applications/Reasonix.app`（または右クリック → 開く → 確認）。
-- **Windows** — SmartScreen が「不明な発行元」と警告します。**詳細情報 → 実行** をクリックしてください。
-- **Linux** — `.deb` と `.AppImage` はそのまま起動でき、追加の手順は不要です。
-
-CLI が引き続き正規のインターフェースです。CLI に入った機能はすべてデスクトップのコンポーザーからも利用できます。
-
-<details>
-<summary><strong>別のフォルダで作業する · chat と code の違い · スキルを書く</strong></summary>
-
-**別のフォルダで作業する。** Reasonix はファイルシステムツールを起動ディレクトリにスコープします。対象を変えるには `--dir` を渡してください。セッション途中での切り替えは設計上サポートされていません（メモリのパスが古いルートと絡まってしまうため）— 一度終了して再起動してください。
-
-~~~bash
-npx reasonix code --dir /path/to/project
-~~~
-
-**`chat` と `code` の選び方。** `code` がデフォルトであり、ファイルシステム / シェルツールと SEARCH/REPLACE レビューを備えた唯一のモードです。`chat` はより軽量でツールをオフにしたシェルです — MCP は接続しつつディスクアクセスのない思考のパートナーが欲しいときに使ってください。
-
-| 得られるもの | `code` | `chat` |
-|---|---|---|
-| ファイルシステムツール + `edit_file` | ✓ | — |
-| SEARCH/REPLACE → `/apply` レビュー | ✓ | — |
-| シェルツール（ゲート付き） | ✓ | — |
-| プランモード · `/todo` · `/skill new` · `/mcp add` | ✓ | — |
-| メモリ（`remember` / `recall_memory`） | プロジェクト + グローバル | グローバルのみ |
-| 設定からの MCP サーバー · ウェブ検索 · `ask_choice` | ✓ | ✓ |
-| コーディング用システムプロンプト | ✓ | 汎用 |
-| セッションのスコープ | ディレクトリごと | 共有デフォルト |
-
-**最初のスキルを書く。** リモートレジストリはありません — 直接書きます。ファイル（`description:` フロントマター + 本文）を編集し、`/skill list` を実行します。本文をインライン展開する代わりに分離したサブエージェントループを生成するには `runAs: subagent` を追加します。
-
-~~~bash
-/skill new my-skill              # <project>/.reasonix/skills/my-skill.md
-/skill new my-skill --global     # プロジェクト横断で使うなら ~/.reasonix/skills
-~~~
-
-**Claude 形式のスキルも読み込まれます。** `<project>/.claude/skills/<name>/SKILL.md` と `~/.claude/skills/` が Reasonix のネイティブパスと並んで読み込まれるため、Claude 形式のスキルを出力するツールはそのまま動きます。例 — 上流アダプターなしで OpenSpec ワークフローを投入する:
-
-~~~bash
-npx openspec init --tools claude    # .claude/skills/openspec-*/SKILL.md を書き出す
-/skill openspec-propose <task>      # その後 Reasonix から呼び出す
-~~~
-
-</details>
-
-<br/>
-
-## 設定
-
-`~/.reasonix/config.json` の1つの JSON ファイルと、`<project>/.reasonix/` 配下のプロジェクトごとのオーバーライドで設定します。すべてのキー、すべてのスラッシュコマンド、スキル / メモリ / フックのオンディスク構造を網羅した完全なバイリンガルリファレンスはこちら:
-
-> 📘 **[設定ガイド](https://esengine.github.io/DeepSeek-Reasonix/configuration.html)** · [中文](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh)
-
-| トピック | 概要 |
-|---|---|
-| [MCP サーバー](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#mcp) | stdio · SSE · Streamable HTTP。1つの仕様形式が `config.json` と `--mcp` の両方で動作。 |
-| [スキル](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#skills) | モデルが呼び出せる Markdown のプレイブック。`inline` または `subagent` モード。 |
-| [メモリ](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#memory) | プレフィックスにピン留めされるユーザー専用の知識。`user` / `feedback` / `project` / `reference` の型。 |
-| [フック](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#hooks) | ライフサイクルイベント時のシェルコマンド。`PreToolUse`（ゲート） · `PostToolUse` · `UserPromptSubmit` · `Stop`。 |
-| [権限](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#permissions) | ワークスペースごとのシェル許可リスト。完全な前方一致。 |
-| [ウェブ検索](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#search) | デフォルトは Mojeek。`/search-engine` でセルフホストの SearXNG や Metaso に切り替え可能。 |
-| [セマンティックインデックス](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#index) | `reasonix index` — ローカルの Ollama または任意の OpenAI 互換の埋め込みエンドポイント。 |
-
-<br/>
-
-## Reasonix の何が違うのか
-
-ループは3つの柱を中心に構成されています。それぞれが、汎用的なエージェントフレームワークには見えてすらいない問題を解決します — それらは別のキャッシュの仕組み向けに設計されているからです。
-
-<sub align="center">
-
-完全なアーキテクチャ解説はこちらから → [Pillar 1 — キャッシュファーストのループ](./docs/ARCHITECTURE.md#pillar-1--cache-first-loop) · [Pillar 2 — ツール呼び出しの修復](./docs/ARCHITECTURE.md#pillar-2--tool-call-repair) · [Pillar 3 — コスト制御](./docs/ARCHITECTURE.md#pillar-3--cost-control-v06)
-
-</sub>
-
-<br/>
-
-## 機能
-
-<p align="center">
-  <img src="docs/assets/feature-grid.svg" alt="Reasonix capabilities — cell-diff renderer, MCP, plan mode, permissions, dashboard, persistent sessions, hooks/skills/memory, semantic search, auto-checkpoints, /effort knob, transcript replay, event log" width="880"/>
-</p>
-
-<br/>
-
-## 比較
-
-|                                   | Reasonix         | Claude Code       | Cursor              | Aider              |
-|-----------------------------------|------------------|-------------------|---------------------|--------------------|
-| バックエンド                       | DeepSeek         | Anthropic         | OpenAI / Anthropic  | 任意（OpenRouter） |
-| ライセンス                         | **MIT**          | クローズド        | クローズド          | Apache 2           |
-| コストプロファイル                 | **タスクあたり低コスト** | プレミアム  | サブスク + 従量     | まちまち           |
-| DeepSeek プレフィックスキャッシュ  | **設計込み**     | 該当なし          | 該当なし            | 偶発的             |
-| 組み込みウェブダッシュボード       | あり             | —                 | 該当なし（IDE）     | —                  |
-| 設定可能なウェブ検索エンジン       | `/search-engine` | —             | —                   | —                  |
-| ワークスペースごとの永続セッション | あり             | 一部              | 該当なし            | —                  |
-| プランモード · MCP · フック · スキル | あり           | あり              | あり                | 一部               |
-| ウェブ検索（Mojeek + SearXNG + Metaso） | あり        | あり              | あり                | あり               |
-| オープンなコミュニティ開発         | あり             | —                 | —                   | あり               |
-
-リアルタイムのキャッシュヒット率、コスト、手法については [`benchmarks/`](./benchmarks/) を参照してください — 数値はモデルの料金とともに変動するため、README ではなくハーネスとともに置いています。
-
-<br/>
-
-## ドキュメント
-
-- [**アーキテクチャ**](./docs/ARCHITECTURE.md) — 3つの柱: キャッシュファーストのループ、ツール呼び出しの修復、コスト制御
-- [**CLI リファレンス**](./docs/CLI-REFERENCE.md) — すべてのシェルサブコマンド、すべてのスラッシュコマンド、すべてのキーバインド
-- [**QQ チャンネルのセットアップ**](./docs/qq-connect.md) — CLI の初回接続フロー、デスクトップのエントリ、QQ Open Platform の認証情報
-- [**ベンチマーク**](./benchmarks/) — τ-bench-lite ハーネス、トランスクリプト、コスト算出手法
-- [**公式サイト**](https://esengine.github.io/DeepSeek-Reasonix/) — はじめに、ダッシュボードのモックアップ、TUI のモックアップ
-- [**コントリビューション**](./CONTRIBUTING.md) — コメントポリシー、エラーハンドリングのルール、自作よりライブラリを優先
-- [**行動規範**](./CODE_OF_CONDUCT.md) · [**セキュリティポリシー**](./SECURITY.md)
-
-<br/>
-
-## コミュニティ
-
-> [!NOTE]
-> Reasonix はオープンソースであり、コミュニティによって開発されています。このファイル末尾の謝辞ウォールにあるすべてのアバターは、実際にマージされた PR です。
-
-スコープが切られた入門用チケット — それぞれ背景、コードの参照ポイント、受け入れ基準、ヒント付き — は [`good first issue`](https://github.com/esengine/reasonix/labels/good%20first%20issue) ラベルの下にあります。オープンなものなら何でも選んでください。
-
-**オープンな Discussions — 意見募集中:**
-
-- [#20 · CLI / TUI デザイン](https://github.com/esengine/reasonix/discussions/20) — 何が壊れていて、何が足りず、何を変えますか？
-- [#21 · ダッシュボードデザイン](https://github.com/esengine/reasonix/discussions/21) — [提案中のモックアップ](https://esengine.github.io/DeepSeek-Reasonix/design/agent-dashboard.html)への反応をどうぞ
-- [#22 · 今後の機能ウィッシュリスト](https://github.com/esengine/reasonix/discussions/22) — 次に Reasonix に組み込みたいものは？
-
-**すでに Reasonix を使っていて、他の人に広める手伝いをしてもいい？** ブログ記事、記事、スクリーンショット、講演、動画を [**Show and tell**](https://github.com/esengine/reasonix/discussions/categories/show-and-tell) に公開してください。このプロジェクトにはマーケティング予算がありません — 新しいユーザーは口コミで見つけてくれます。継続的に応援してくれる人は、以下のバッジを獲得でき、付与されるとコントリビューターウォールの横に表示されます:
-
-<p align="center">
-  <a href="https://github.com/esengine/reasonix/discussions/categories/show-and-tell">
-    <img src="https://img.shields.io/badge/REASONIX-📣%20ADVOCATE-c4b5fd?style=for-the-badge&labelColor=0d1117" alt="Reasonix Advocate badge — earned by sustained advocates"/>
-  </a>
-</p>
-
-**最初の PR の前に**: [`CONTRIBUTING.md`](./CONTRIBUTING.md) を読んでください — 短く厳格なルール（コメント、エラー、自作よりライブラリを優先）。`tests/comment-policy.test.ts` がコメントのルールを強制します。`npm run verify` がプッシュ前のゲートです。参加することで [行動規範](./CODE_OF_CONDUCT.md) に同意したものとみなされます。セキュリティの問題は → [SECURITY.md](./SECURITY.md)。
-
-<br/>
-
-## 非目標
-
-> [!IMPORTANT]
-> Reasonix には明確な方針があります。あえて *やらない* ことがいくつかあります — あなたの作業に合ったツールを選べるよう、ここに挙げておきます。
-
-- **マルチプロバイダーの柔軟性。** あえて DeepSeek 専用にしています。1つのバックエンドに密結合していることは制限ではなく機能です。
-- **IDE 統合。** ターミナルファースト。差分は `git diff` に、ファイルツリーは `ls` にあります。ダッシュボードはあくまで補助であり、Cursor の代替ではありません。
-- **最難関リーダーボードの推論。** Claude Opus はいまだに一部のベンチマークで勝っています。DeepSeek はコーディングで競争力がありますが、もしあなたの作業が「この認証バグを直す」ではなく「この博士課程レベルの証明を解く」ものなら、Claude から始めてください。
-- **エアギャップ / 完全無料。** Reasonix には有料の DeepSeek API キーが必要です。エアギャップ環境やゼロコストでの実行には Aider + Ollama または [Continue](https://continue.dev) を参照してください。
-
-<br/>
-
-## スター履歴
-
-<a href="https://www.star-history.com/?repos=esengine%2FDeepSeek-Reasonix&type=date&legend=top-left">
- <picture>
-   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=esengine/DeepSeek-Reasonix&type=date&theme=dark&legend=top-left" />
-   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=esengine/DeepSeek-Reasonix&type=date&legend=top-left" />
-   <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=esengine/DeepSeek-Reasonix&type=date&legend=top-left" />
- </picture>
-</a>
-
-<br/>
-
-## サポート
-
-Reasonix が役に立って、お礼を言いたいと思ったらどうぞ。これはあくまでコーヒー一杯であって契約ではありません — 寄付で機能の優先順位が買えるわけでも、issue のトリアージのされ方が変わるわけでもありません。
-
-- **International** — PayPal: [paypal.me/yuhuahui](https://paypal.me/yuhuahui)
-- **国内** — 微信支付（扫码）
-
-<p align="center">
-  <img src=".github/sponsor/wechat-pay.jpg" alt="WeChat Pay QR code" width="240"/>
-</p>
-
-<br/>
-
-## 謝辞
-
-Reasonix を最も形づくってくれた人々の小さなリストです — コミット数とコード量の両方で測っています。**重要度の順位付けはなく、アルファベット順に記載しています。** 完全なコントリビューターグラフは
-[GitHub](https://github.com/esengine/DeepSeek-Reasonix/graphs/contributors) にあります。
-
-- [**ctharvey**](https://github.com/ctharvey)
-- [**dimasd-angga**](https://github.com/dimasd-angga) (Dimas D. Angga)
-- [**Evan-Pycraft**](https://github.com/Evan-Pycraft)
-- [**ForeverYoungPp**](https://github.com/ForeverYoungPp)
-- [**GTC2080**](https://github.com/GTC2080) (TaoMu)
-- [**kabaka9527**](https://github.com/kabaka9527)
-- [**lisniuse**](https://github.com/lisniuse) (Richie)
-- [**wade19990814-hue**](https://github.com/wade19990814-hue)
-- [**wviana**](https://github.com/wviana) (Wesley Viana)
-
-また、プロジェクトのロゴをデザインしてくれた [**Bernardxu123**](https://github.com/Bernardxu123)（[`docs/brand/`](./docs/brand/) を参照）と、小紅書（XiaoHongShu）でプロジェクトを宣伝してくれた [AIGC Link](https://xhslink.com/m/80ngts127cA) にも別途感謝します。
-
-<p align="center">
-  <a href="https://github.com/esengine/DeepSeek-Reasonix/graphs/contributors">
-    <img src="https://contrib.rocks/image?repo=esengine/DeepSeek-Reasonix&max=100&columns=12" alt="Contributors to esengine/DeepSeek-Reasonix" width="860"/>
-  </a>
-</p>
-
-<br/>
-
----
-
-<p align="center">
-  <sub>MIT — <a href="./LICENSE">LICENSE</a> を参照</sub>
-  <br/>
-  <sub><a href="https://github.com/esengine/DeepSeek-Reasonix/graphs/contributors">esengine/DeepSeek-Reasonix</a> のコミュニティによって構築</sub>
-</p>

README.md

@@ -2,35 +2,28 @@
   <img src="docs/logo.svg" alt="Reasonix" width="640"/>
 </p>
 
-> [!WARNING]
-> **This is the legacy TypeScript line (Reasonix 0.x), in maintenance mode.**
-> Active development moved to the **Go rewrite** — see the [`main-v2`](https://github.com/esengine/DeepSeek-Reasonix/tree/main-v2) branch (the new default) and the [migration guide](https://github.com/esengine/DeepSeek-Reasonix/blob/main-v2/docs/MIGRATING.md). Only bug fixes land here.
-
-
 <p align="center">
   <strong>English</strong>
   &nbsp;·&nbsp;
   <a href="./README.zh-CN.md">简体中文</a>
   &nbsp;·&nbsp;
-  <a href="./README.ja-JP.md">日本語</a>
+  <a href="./docs/SPEC.md">Spec</a>
   &nbsp;·&nbsp;
   <a href="https://esengine.github.io/DeepSeek-Reasonix/">Website</a>
   &nbsp;·&nbsp;
-  <a href="https://esengine.github.io/DeepSeek-Reasonix/configuration.html">Guide</a>
-  &nbsp;·&nbsp;
-  <a href="./docs/ARCHITECTURE.md">Architecture</a>
-  &nbsp;·&nbsp;
-  <a href="./benchmarks/">Benchmarks</a>
-  &nbsp;·&nbsp;
   <strong><a href="https://discord.gg/XF78rEME2D">Discord</a></strong>
 </p>
 
+> [!IMPORTANT]
+> **Reasonix 1.0 is a ground-up rewrite in Go** — this branch (`main-v2`) is the new default and where development happens now.
+> The earlier `0.x` TypeScript releases are **legacy**, living on the [`v1`](https://github.com/esengine/DeepSeek-Reasonix/tree/v1) branch (maintenance only).
+> See the **[migration guide](./docs/MIGRATING.md)**. `npm i -g reasonix` stays the install command — `1.0.0`+ delivers the Go binary, `0.x` is the legacy TS build.
+
 <p align="center">
   <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/v/reasonix.svg?style=flat-square&color=cb3837&labelColor=161b22&logo=npm&logoColor=white" alt="npm version"/></a>
   <a href="https://github.com/esengine/reasonix/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/esengine/reasonix/ci.yml?style=flat-square&label=ci&labelColor=161b22&logo=githubactions&logoColor=white" alt="CI"/></a>
   <a href="./LICENSE"><img src="https://img.shields.io/npm/l/reasonix.svg?style=flat-square&color=8b949e&labelColor=161b22" alt="license"/></a>
   <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/dm/reasonix.svg?style=flat-square&color=3fb950&labelColor=161b22&label=downloads" alt="downloads"/></a>
-  <a href="./package.json"><img src="https://img.shields.io/node/v/reasonix.svg?style=flat-square&color=5fa04e&labelColor=161b22&logo=nodedotjs&logoColor=white" alt="node"/></a>
   <a href="https://github.com/esengine/reasonix/stargazers"><img src="https://img.shields.io/github/stars/esengine/reasonix.svg?style=flat-square&color=dbab09&labelColor=161b22&logo=github&logoColor=white" alt="GitHub stars"/></a>
   <a href="https://atomgit.com/esengine/DeepSeek-Reasonix"><img src="https://atomgit.com/esengine/DeepSeek-Reasonix/star/badge.svg" alt="AtomGit stars"/></a>
   <a href="https://github.com/esengine/reasonix/graphs/contributors"><img src="https://img.shields.io/github/contributors/esengine/reasonix.svg?style=flat-square&color=bc8cff&labelColor=161b22&logo=github&logoColor=white" alt="contributors"/></a>
@@ -47,236 +40,257 @@
 <br/>
 
 <h3 align="center">A DeepSeek-native AI coding agent for your terminal.</h3>
-<p align="center">Engineered around prefix-cache stability — so token costs stay low across long sessions, and you can leave it running.</p>
+<p align="center">A config- and plugin-driven harness — a single static Go binary, tuned around DeepSeek's prefix cache so token costs stay low across long sessions.</p>
 
 <br/>
 
-<p align="center">
-  <img src="docs/assets/hero-terminal.svg" alt="Reasonix code mode — assistant proposes a SEARCH/REPLACE edit; nothing on disk until /apply" width="860"/>
-</p>
-
-<br/>
-
-> [!TIP]
-> **Cache stability isn't a feature you turn on; it's an invariant the loop is designed around.** That's the whole reason Reasonix is DeepSeek-only — every layer is tuned to the byte-stable prefix-cache mechanic.
-
-> [!NOTE]
-> **Real user, single day (2026-05-01):** 435M input tokens, **99.82% cache hit**, ~$12 instead of the ~$61 the same workload would cost with no cache on `v4-flash` — see the [case study](./benchmarks/real-world-cache/README.md). DeepSeek provides the cacheable bytes; the four mechanisms in [Pillar 1](./docs/ARCHITECTURE.md#pillar-1--cache-first-loop) are how Reasonix keeps them cacheable across long sessions.
-
 > [!IMPORTANT]
-> **Community · 加入社区** — bilingual Discord with channels for setup help (`#help` / `#求助`), workflow showcases, feature ideas, and contributor-only PR coordination. Verify your GitHub in-server to get the **Contributor** role automatically. → **<https://discord.gg/XF78rEME2D>**
+> **Community · 加入社区** — bilingual Discord for setup help (`#help` / `#求助`), workflow showcases, and feature ideas. → **<https://discord.gg/XF78rEME2D>**
 
 <br/>
 
+## Features
+
+- **Config-driven.** Providers, the agent, enabled tools, and plugins are all
+  declared in `reasonix.toml`. No hardcoded models.
+- **Multi-model & composable.** DeepSeek (flash/pro) and MiMo ship as presets;
+  any OpenAI-compatible endpoint is a config entry, not new code. Optionally run
+  two models together (executor + planner) in separate, cache-stable sessions.
+- **Plugin-driven.** External tools run as subprocesses over stdio JSON-RPC
+  (MCP-compatible). Built-in tools self-register at compile time.
+- **Zero-friction distribution.** `CGO_ENABLED=0` single binary; cross-compile
+  to six targets with one command. The only dependency is a TOML parser.
+
 ## Install
 
-Requires Node ≥ 22. Works on macOS · Linux · Windows (PowerShell · Git Bash · Windows Terminal).
+```sh
+npm i -g reasonix                  # any OS; pulls the prebuilt native binary
+brew install esengine/reasonix/reasonix   # macOS
+```
 
-Install Reasonix globally if you want the `reasonix` command available on your `PATH`:
+Prebuilt archives (`darwin|linux|windows × amd64|arm64`) and `SHA256SUMS` are on
+every [GitHub release](https://github.com/esengine/DeepSeek-Reasonix/releases).
 
-~~~bash
-npm install -g reasonix
-reasonix code my-project   # paste a DeepSeek API key on first run; persists after
-~~~
+### Build from source
 
-Or run it once without installing globally:
+```sh
+make build      # -> bin/reasonix(.exe)
+make cross      # -> dist/ (darwin|linux|windows × amd64|arm64)
+```
 
-~~~bash
-cd my-project
-npx reasonix code          # always uses the latest package by default
-~~~
+## Quick start
 
-Grab a [DeepSeek API key →](https://platform.deepseek.com/api_keys) · `reasonix code --help` for flags.
-
-If you use Reasonix daily, global install is the simplest path. If you just want to try it, use `npx`.
-
-**Prefer fewer keystrokes?** The shorter `dsnix` alias resolves to the same CLI:
-
-~~~bash
-npm install -g dsnix       # exposes `dsnix` on PATH, depends on reasonix
-npx dsnix@latest code      # one-shot via the shorter command
-~~~
-
-A global `npm install -g reasonix` also drops a `dsnix` shim on PATH, so the two are interchangeable.
-
-Bare `reasonix` (no subcommand) launches `code` in the current directory — typing `reasonix` and `reasonix code` are equivalent.
-
-| Command | When |
-|---|---|
-| `reasonix` / `reasonix code [dir]` | The coding agent. **Start here.** |
-| `reasonix chat` | Plain chat — no filesystem or shell tools. |
-| `reasonix run "task"` | One-shot, streams to stdout. Good for pipes. |
-| `reasonix doctor` | Health check: Node, API key, MCP wiring. |
-| `reasonix update` | Upgrade Reasonix itself. |
-
-Other subcommands (`replay` · `diff` · `events` · `stats` · `index` · `mcp` · `prune-sessions`) are in `reasonix --help` and the [CLI reference](https://esengine.github.io/DeepSeek-Reasonix/#cli).
-
-### QQ channel
-
-QQ can extend an existing `chat`, `code`, or desktop session as a remote channel. It is part of the current session flow, not a separate runtime mode.
-
-- CLI: start a session, then run `/qq connect`
-- Desktop: open `Settings -> General -> QQ Channel`
-
-Once connected, QQ messages can enter the current session, assistant replies route back to QQ, and follow-up interactions can continue remotely.
-
-For full setup, desktop quick start, and troubleshooting, see [QQ channel setup](./docs/qq-connect.md).
-
-### Desktop client (prerelease)
-
-A native Tauri client for users who want a GUI over the same loop. Multi-tab, the right-panel shows files the agent has read or edited this session, the same cost / cache / token meters live at the bottom. Same DeepSeek API key, same `~/.reasonix` config — the desktop bundles its own Node runtime, no separate `npm install` step.
-
-Download platform installers from [GitHub Releases](https://github.com/esengine/DeepSeek-Reasonix/releases). The desktop ships as a **prerelease**: the loop and protocol are the same as the CLI, but the UI is still being polished and the installers aren't code-signed yet.
-
-- **macOS** — first launch hits Gatekeeper. One-time fix: `xattr -dr com.apple.quarantine /Applications/Reasonix.app` (or right-click → Open → confirm).
-- **Windows** — SmartScreen warns "Unknown publisher". Click **More info → Run anyway**.
-- **Linux** — `.deb` and `.AppImage` ship plain, no extra step.
-
-The CLI remains the canonical surface. Anything that lands in the CLI is also available from the desktop's composer.
-
-<details>
-<summary><strong>Working in another folder · chat vs. code · author a skill</strong></summary>
-
-**Working in a different folder.** Reasonix scopes filesystem tools to the launch directory; pass `--dir` to retarget. Mid-session switching isn't supported by design (memory paths would tangle with stale roots) — quit and relaunch.
-
-~~~bash
-npx reasonix code --dir /path/to/project
-~~~
-
-**Picking `chat` vs `code`.** `code` is the default and the only mode with filesystem / shell tools and SEARCH/REPLACE review. `chat` is the lighter, tools-off shell — reach for it when you want a thinking partner with MCP attached but no disk access.
-
-| What you get | `code` | `chat` |
-|---|---|---|
-| Filesystem tools + `edit_file` | ✓ | — |
-| SEARCH/REPLACE → `/apply` review | ✓ | — |
-| Shell tool (gated) | ✓ | — |
-| Plan mode · `/todo` · `/skill new` · `/mcp add` | ✓ | — |
-| Memory (`remember` / `recall_memory`) | project + global | global only |
-| MCP servers from config · web search · `ask_choice` | ✓ | ✓ |
-| Coding system prompt | ✓ | generic |
-| Session scope | per-directory | shared default |
-
-**Author your first skill.** No remote registry — write them directly. Edit the file (`description:` frontmatter + body), then `/skill list`. Add `runAs: subagent` to spawn an isolated subagent loop instead of inlining the body.
-
-~~~bash
-/skill new my-skill              # <project>/.reasonix/skills/my-skill.md
-/skill new my-skill --global     # ~/.reasonix/skills for cross-project use
-~~~
-
-**Claude-format skills also load.** `<project>/.claude/skills/<name>/SKILL.md` and `~/.claude/skills/` are read alongside Reasonix's native paths, so tooling that emits Claude-format skills works out of the box. Example — drop OpenSpec workflows in without an upstream adapter:
-
-~~~bash
-npx openspec init --tools claude    # writes .claude/skills/openspec-*/SKILL.md
-/skill openspec-propose <task>      # then invoke from Reasonix
-~~~
-
-</details>
-
-<br/>
+```sh
+reasonix setup                      # config wizard → ./reasonix.toml
+export DEEPSEEK_API_KEY=sk-...  # or put it in .env (see .env.example)
+reasonix chat                       # then run /init to generate AGENTS.md (project memory)
+reasonix run "implement the TODOs in main.go"
+reasonix run --model mimo-pro "add unit tests for this function"
+echo "explain this code" | reasonix run
+```
 
 ## Configuration
 
-One JSON file at `~/.reasonix/config.json` plus per-project overrides under `<project>/.reasonix/`. The full bilingual reference — every key, every slash command, the on-disk shape of skills/memory/hooks — lives at:
+Resolution order: **flag > `./reasonix.toml` > `~/.config/reasonix/config.toml` >
+built-in defaults**. Secrets come from the environment via `api_key_env` and are
+never stored in config files.
 
-> 📘 **[Configuration Guide](https://esengine.github.io/DeepSeek-Reasonix/configuration.html)** · [中文](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh)
+```toml
+default_model = "deepseek-flash"   # executor; set [agent].planner_model to add a planner
+# language    = "zh"               # ui language; empty = auto-detect from $LANG / $REASONIX_LANG
 
-| Topic | Quick read |
-|---|---|
-| [MCP servers](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#mcp) | stdio · SSE · Streamable HTTP. One spec format works for both `config.json` and `--mcp`. |
-| [Skills](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#skills) | Markdown playbooks the model can invoke. `inline` or `subagent` mode. |
-| [Memory](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#memory) | User-private knowledge pinned into the prefix. `user` / `feedback` / `project` / `reference` types. |
-| [Hooks](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#hooks) | Shell commands on lifecycle events. `PreToolUse` (gating) · `PostToolUse` · `UserPromptSubmit` · `Stop`. |
-| [Permissions](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#permissions) | Per-workspace shell allowlist. Exact-prefix match. |
-| [Web search](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#search) | Bing by default; switch to Baidu AI Search, self-hosted SearXNG, Metaso, Tavily, Perplexity, Exa, Brave, or Ollama with `/search-engine`. |
-| [Semantic index](https://esengine.github.io/DeepSeek-Reasonix/configuration.html#index) | `reasonix index` — local Ollama or any OpenAI-compatible embedding endpoint. |
+[agent]
+# planner_model = "mimo-pro"          # optional low-frequency planner
+# subagent_model = "deepseek-pro"     # optional default for runAs=subagent skills
+# subagent_models = { review = "deepseek-pro", security_review = "deepseek-pro" }
+auto_plan = "off"                  # off|on; off keeps plan mode manual
+# auto_plan_classifier = "deepseek-flash"   # optional; only borderline tasks call it
 
-<br/>
+[[providers]]
+name        = "deepseek-flash"
+kind        = "openai"
+base_url    = "https://api.deepseek.com"
+model       = "deepseek-v4-flash"
+api_key_env = "DEEPSEEK_API_KEY"
+# also preset: deepseek-pro, mimo-pro (mimo-v2.5-pro), mimo-flash (mimo-v2-flash) @ api.xiaomimimo.com/v1
 
-## What makes Reasonix different
+[tools]
+enabled = []   # omit/empty = all built-ins
 
-The loop is organized around three pillars. Each one solves a problem generic agent frameworks don't even see — because they were designed for a different cache mechanic.
+[skills]
+# paths = ["~/my-skills", "../shared/skills"]   # extra custom skill roots
+# disabled_skills = ["review"]                  # hide skills until /skill enable <name>
 
-<sub align="center">
+[permissions]
+mode  = "ask"                                # writer fallback when no rule matches: ask|allow|deny
+deny  = ["bash(rm -rf*)", "bash(git push*)"] # hard-blocked in every mode
+allow = ["bash(go test*)"]                   # never prompted
 
-Click through to the full architecture writeup → [Pillar 1 — Cache-first loop](./docs/ARCHITECTURE.md#pillar-1--cache-first-loop) · [Pillar 2 — Tool-call repair](./docs/ARCHITECTURE.md#pillar-2--tool-call-repair) · [Pillar 3 — Cost control](./docs/ARCHITECTURE.md#pillar-3--cost-control-v06)
+[sandbox]
+# workspace_root = ""          # file-writers confined here; empty = current dir
+# allow_write    = ["/tmp"]    # extra dirs write_file/edit_file/multi_edit may touch
 
-</sub>
+[[plugins]]
+name    = "example"
+command = "reasonix-plugin-example"
+```
 
-<br/>
+Permissions gate each tool call: `deny` > `ask` > `allow` > fallback (readers
+always allow; writers fall back to `mode`). `reasonix chat` prompts before writers
+(`y` once · `a` this session · `n` no); `reasonix run` stays autonomous but still
+honours `deny`. See [`docs/SPEC.md`](docs/SPEC.md) for the full schema and contract.
 
-## Capabilities
+Permissions are *policy* (which calls to allow / prompt). The **sandbox** is
+*enforcement*: the file-writers (`write_file` / `edit_file` / `multi_edit`)
+refuse any path outside `[sandbox] workspace_root` (default: the current dir, so
+edits stay in the project), resolving symlinks and `..` so a link can't tunnel
+out. Reads are unrestricted. `bash` is itself jailed on macOS by default
+(`[sandbox] bash`, Seatbelt): commands may write only those same roots (plus
+temp and toolchain caches) and reach the network only when `[sandbox] network`
+is set. Other platforms fall back to running unconfined for now (see
+`docs/SPEC.md` §9 for the escape-prompt and Linux support still to come).
 
-<p align="center">
-  <img src="docs/assets/feature-grid.svg" alt="Reasonix capabilities — cell-diff renderer, MCP, plan mode, permissions, dashboard, persistent sessions, hooks/skills/memory, semantic search, auto-checkpoints, /effort knob, transcript replay, event log" width="880"/>
-</p>
+### Plugins (MCP)
 
-<br/>
+Reasonix is an MCP client. A `[[plugins]]` entry's `type` selects the transport:
+`stdio` (default) launches a local subprocess (`command`/`args`/`env`); `http`
+(Streamable HTTP) connects to a remote `url` with optional static `headers`
+(`${VAR}` / `${VAR:-default}` expanded from the environment, so tokens stay out
+of the file). Tools surface to the model as `mcp__<server>__<tool>`; a tool
+declaring MCP's `readOnlyHint: true` joins parallel dispatch and the permission
+reader-default.
 
-## How it compares
+A server's **prompts** surface as `/mcp__<server>__<prompt>` slash commands
+(positional args after the command); its **resources** are pulled in by writing
+`@<server>:<uri>` in a message; `/mcp` lists connected servers and what each
+exposes. `make build` also produces `bin/reasonix-plugin-example` — a runnable
+reference stdio server (`echo`, `wordcount`, a `review` prompt, a style-guide
+resource) you can copy.
 
-|                                   | Reasonix         | Claude Code       | Cursor              | Aider              |
-|-----------------------------------|------------------|-------------------|---------------------|--------------------|
-| Backend                           | DeepSeek         | Anthropic         | OpenAI / Anthropic  | any (OpenRouter)   |
-| License                           | **MIT**          | closed            | closed              | Apache 2           |
-| Cost profile                      | **low per task** | premium           | subscription + use  | varies             |
-| DeepSeek prefix-cache             | **engineered**   | not applicable    | not applicable      | incidental         |
-| Embedded web dashboard            | yes              | —                 | n/a (IDE)           | —                  |
-| Configurable web search engine    | `/search-engine` | —             | —                   | —                  |
-| Persistent per-workspace sessions | yes              | partial           | n/a                 | —                  |
-| Plan mode · MCP · hooks · skills  | yes              | yes               | yes                 | partial            |
-| Web search (Bing + Baidu + SearXNG + API engines) | yes              | yes               | yes                 | yes                |
-| Open community development        | yes              | —                 | —                   | yes                |
+```toml
+[[plugins]]                       # local stdio server
+name    = "example"
+command = "reasonix-plugin-example"
 
-For live cache-hit rates, costs, and methodology, see [`benchmarks/`](./benchmarks/) — the numbers move with model pricing, so they live with the harness, not in the README.
+[[plugins]]                       # remote server over Streamable HTTP
+name    = "stripe"
+type    = "http"
+url     = "https://mcp.stripe.com"
+headers = { Authorization = "Bearer ${STRIPE_KEY}" }
+```
 
-<br/>
+**Already have an `.mcp.json`?** Drop it in the project root and Reasonix
+reads it as-is — the `mcpServers` spec (`command`/`args`/`env`, `type`/`url`/
+`headers`, `${VAR}` expansion) maps field-for-field onto `[[plugins]]`. Both
+sources are merged; on a name collision `reasonix.toml` wins.
 
-## Documentation
+```json
+{
+  "mcpServers": {
+    "filesystem": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path"] },
+    "stripe": { "type": "http", "url": "https://mcp.stripe.com", "headers": { "Authorization": "Bearer ${STRIPE_KEY}" } }
+  }
+}
+```
 
-- [**Architecture**](./docs/ARCHITECTURE.md) — three pillars: cache-first loop, tool-call repair, cost control
-- [**CLI Reference**](./docs/CLI-REFERENCE.md) — every shell subcommand, every slash command, every keybinding
-- [**QQ channel setup**](./docs/qq-connect.md) — CLI first-connect flow, desktop entry, and QQ Open Platform credentials
-- [**Benchmarks**](./benchmarks/) — τ-bench-lite harness, transcripts, cost methodology
-- [**Website**](https://esengine.github.io/DeepSeek-Reasonix/) — getting started, dashboard mockup, TUI mockup
-- [**Contributing**](./CONTRIBUTING.md) — comment policy, error-handling rules, library-over-hand-rolled
-- [**Code of Conduct**](./CODE_OF_CONDUCT.md) · [**Security policy**](./SECURITY.md)
+**Upgrading from `0.x`?** Your old `~/.reasonix/config.json` is still read for its
+`mcpServers` (honouring `mcpDisabled`) as a lowest-priority source, so MCP servers
+keep working — move them into `reasonix.toml`'s `[[plugins]]` or a `.mcp.json` when
+convenient.
 
-<br/>
+### Slash commands
 
-## Community
+In `reasonix chat`, built-in commands (`/compact`, `/new`, `/rewind`, `/tree`,
+`/branch`, `/switch`, `/todo`, `/model`, `/effort`, `/mcp`, `/memory`, `/help`) run locally.
+`/tree` shows saved conversation branches, `/branch [name]` forks the current
+conversation tip, `/branch <turn> [name]` forks from an earlier checkpointed turn,
+and `/switch <id|name>` loads another branch. **Custom commands** are Markdown files under
+`.reasonix/commands/` (project) or `~/.config/reasonix/commands/` (user) —
+`review.md` becomes `/review`, a subdirectory namespaces it (`git/commit.md` →
+`/git:commit`). The body is a prompt template; invoking the command sends it as a
+turn.
 
-> [!NOTE]
-> Reasonix is open source and community-developed. Every avatar in the Acknowledgments wall at the bottom of this file is a real PR that shipped.
+```markdown
+---
+description: Review the staged diff
+argument-hint: [focus-area]
+---
+Review the staged diff. Focus on $ARGUMENTS, list bugs with file:line.
+```
 
-Scoped starter tickets — each with background, code pointers, acceptance criteria, and hints — live under the [`good first issue`](https://github.com/esengine/reasonix/labels/good%20first%20issue) label. Pick anything open.
+`$ARGUMENTS` expands to all space-separated args, `$1`…`$N` to positional ones.
+MCP prompts also appear here as `/mcp__<server>__<prompt>`.
 
-**Open Discussions — opinions wanted:**
+### @ references
 
-- [#20 · CLI / TUI design](https://github.com/esengine/reasonix/discussions/20) — what's broken, what's missing, what would you change?
-- [#21 · Dashboard design](https://github.com/esengine/reasonix/discussions/21) — react against the [proposed mockup](https://esengine.github.io/DeepSeek-Reasonix/design/agent-dashboard.html)
-- [#22 · Future feature wishlist](https://github.com/esengine/reasonix/discussions/22) — what would you build into Reasonix next?
+Embed `@` references in a message and Reasonix resolves them before sending, as
+tagged context blocks: `@path/to/file` (or `@dir`) injects a local file's
+contents (or a directory listing), and `@<server>:<uri>` injects an MCP
+resource. A local path is only treated as a reference when it actually exists,
+so ordinary `@mentions` stay literal. Typing `/` or `@` opens an autocomplete
+menu — slash commands, or hierarchical file navigation (one directory level at a
+time, descend into folders) plus MCP resources.
 
-**Already using Reasonix and willing to help others discover it?** Publish blog posts, articles, screenshots, talks, or videos to [**Show and tell**](https://github.com/esengine/reasonix/discussions/categories/show-and-tell). The project has no marketing budget — community word of mouth is how new users find it. Sustained advocates earn the badge below, displayed next to the contributors wall once awarded:
+### Two-model collaboration (optional)
 
-<p align="center">
-  <a href="https://github.com/esengine/reasonix/discussions/categories/show-and-tell">
-    <img src="https://img.shields.io/badge/REASONIX-📣%20ADVOCATE-c4b5fd?style=for-the-badge&labelColor=0d1117" alt="Reasonix Advocate badge — earned by sustained advocates"/>
-  </a>
-</p>
+`reasonix setup` keeps first-run minimal: pick provider → keys (every SKU of a
+chosen provider is enabled). Running two models together (executor + planner,
+separate cache-stable sessions) is a one-line edit afterwards — set
+`planner_model` to any other enabled provider:
 
-**Before your first PR**: read [`CONTRIBUTING.md`](./CONTRIBUTING.md) — short, strict rules (comments, errors, libraries-over-hand-rolled). `tests/comment-policy.test.ts` enforces the comment ones; `npm run verify` is the pre-push gate. By participating you agree to the [Code of Conduct](./CODE_OF_CONDUCT.md). Security issues → [SECURITY.md](./SECURITY.md).
+```toml
+[agent]
+planner_model = "deepseek-pro"   # used as the low-frequency planner
+```
 
-<br/>
+Subagent skills inherit the executor model by default. Set `subagent_model` to
+run them on another configured model, or use `subagent_models` to override only
+specific skills such as `review` or `security_review`.
 
-## Non-goals
+For interactive frontends, plan mode is manual by default. Set
+`agent.auto_plan = "on"` to make complex-looking tasks enter plan mode
+automatically: Reasonix first drafts a read-only plan, then waits for approval
+before editing or running side-effecting commands. `auto_plan_classifier` can
+name a cheap provider such as `deepseek-flash`; it is only called for borderline
+inputs and falls back to the heuristic if classification fails. Use
+`/auto-plan off|on` in `reasonix chat` to change the user-level setting, or
+`reasonix config auto-plan off|on` from a shell/script. Pass `--local` to the
+shell command only when you intentionally want a project-local override.
 
-> [!IMPORTANT]
-> Reasonix is opinionated. Some things it deliberately *doesn't* do — listed here so you can pick the right tool for your work.
+## Architecture
 
-- **Multi-provider flexibility.** DeepSeek-only on purpose. Coupling to one backend is the feature, not a limitation.
-- **IDE integration.** Terminal-first. The diff lives in `git diff`, the file tree in `ls`. The dashboard is a companion, not a Cursor replacement.
-- **Hardest-leaderboard reasoning.** Claude Opus still wins some benchmarks. DeepSeek is competitive on coding; if your work is "solve this PhD proof" rather than "fix this auth bug," start with Claude.
-- **Air-gapped / fully-free.** Reasonix needs a paid DeepSeek API key. For air-gapped or zero-cost runs see Aider + Ollama or [Continue](https://continue.dev).
+Three tiers of extensibility, all behind registries the core resolves by name:
+
+1. **Registry** — `Provider` and `Tool` are interfaces; the core has no
+   `switch model`.
+2. **Compile-time built-ins** — providers (`provider/openai`) and tools
+   (`tool/builtin`) self-register via `init()`; `main` blank-imports them.
+   Adding a built-in is one file plus one import.
+3. **Runtime plugins** — executables declared in config, spoken to over
+   newline-delimited JSON-RPC 2.0 on stdin/stdout (the MCP stdio convention).
+   Each remote tool is adapted to the `Tool` interface.
+
+## Status
+
+Done: registry-based providers/tools, OpenAI-compatible streaming with tool
+calls (bounded retry on 429/5xx), built-in tools (read_file, write_file,
+edit_file, multi_edit, bash, ls, glob, grep, web_fetch, task, todo_write, ask),
+TOML config, an interactive `reasonix setup` wizard, two-model collaboration
+(executor + planner in separate, cache-stable sessions), low-frequency context
+compaction, sub-agents (`task`), a bubbletea chat TUI (markdown, plan mode with
+controller-driven approval, live token/activity readout, pinned task list,
+`ask` question chooser, `/compact` `/new` `/tree` `/branch` `/switch` `/todo`), session persistence + resume,
+per-call **permissions** (allow/ask/deny rules; chat prompts before writers, deny
+rules hard-block everywhere), a **workspace sandbox** confining file-writers to
+the project (symlink/`..`-safe), an MCP client — **stdio + Streamable HTTP**
+transports, tools (`mcp__server__tool`, `readOnlyHint`-aware), prompts (slash
+commands), resources (`@`-references), and `/mcp`, configured via `[[plugins]]`
+or a project `.mcp.json` — custom slash commands (`.reasonix/commands/*.md`),
+`@file` / `@resource` references, plus a runnable reference plugin
+(`cmd/reasonix-plugin-example`), the harness loop, and CLI. A Wails desktop
+client (`desktop/`) drives the same kernel. Next: an OS-level sandbox for `bash`
+(macOS Seatbelt / Linux bubblewrap), an Anthropic-native provider, MCP OAuth +
+legacy SSE. See `docs/SPEC.md` §9.
 
 <br/>
 
@@ -323,7 +337,7 @@ of importance.** The full contributor graph is on
 - [**wviana**](https://github.com/wviana) (Wesley Viana)
 
 Also a separate thank-you to [**Bernardxu123**](https://github.com/Bernardxu123)
-for designing the project logo (see [`docs/brand/`](./docs/brand/)), and to
+for designing the project logo, and to
 [AIGC Link](https://xhslink.com/m/80ngts127cA) for promoting the project on XiaoHongShu.
 
 <p align="center">

README.zh-CN.md

@@ -2,35 +2,28 @@
   <img src="docs/logo.svg" alt="Reasonix" width="640"/>
 </p>
 
-> [!WARNING]
-> **这是 legacy TypeScript 线(Reasonix 0.x),仅维护。**
-> 开发已转移到 **Go 重写版** —— 见 [`main-v2`](https://github.com/esengine/DeepSeek-Reasonix/tree/main-v2) 分支(新的默认分支)与[迁移指南](https://github.com/esengine/DeepSeek-Reasonix/blob/main-v2/docs/MIGRATING.md)。此处仅修复 bug。
-
-
 <p align="center">
   <a href="./README.md">English</a>
   &nbsp;·&nbsp;
   <strong>简体中文</strong>
   &nbsp;·&nbsp;
-  <a href="./README.ja-JP.md">日本語</a>
+  <a href="./docs/SPEC.md">规格</a>
   &nbsp;·&nbsp;
   <a href="https://esengine.github.io/DeepSeek-Reasonix/">官方网站</a>
   &nbsp;·&nbsp;
-  <a href="https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh">配置指南</a>
-  &nbsp;·&nbsp;
-  <a href="./docs/ARCHITECTURE.md">架构文档</a>
-  &nbsp;·&nbsp;
-  <a href="./benchmarks/">基准测试</a>
-  &nbsp;·&nbsp;
   <strong><a href="https://discord.gg/XF78rEME2D">Discord</a></strong>
 </p>
 
+> [!IMPORTANT]
+> **Reasonix 1.0 是用 Go 从零重写的版本** —— 本分支(`main-v2`)已是新的默认分支,后续开发都在这里。
+> 早期的 `0.x` TypeScript 版本转为 **legacy**,保留在 [`v1`](https://github.com/esengine/DeepSeek-Reasonix/tree/v1) 分支(仅维护)。
+> 详见**[迁移指南](./docs/MIGRATING.md)**。`npm i -g reasonix` 仍是安装命令——`1.0.0`+ 装的是 Go 二进制,`0.x` 是 legacy TS 版。
+
 <p align="center">
   <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/v/reasonix.svg?style=flat-square&color=cb3837&labelColor=161b22&logo=npm&logoColor=white" alt="npm version"/></a>
   <a href="https://github.com/esengine/reasonix/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/esengine/reasonix/ci.yml?style=flat-square&label=ci&labelColor=161b22&logo=githubactions&logoColor=white" alt="CI"/></a>
   <a href="./LICENSE"><img src="https://img.shields.io/npm/l/reasonix.svg?style=flat-square&color=8b949e&labelColor=161b22" alt="license"/></a>
   <a href="https://www.npmjs.com/package/reasonix"><img src="https://img.shields.io/npm/dm/reasonix.svg?style=flat-square&color=3fb950&labelColor=161b22&label=downloads" alt="downloads"/></a>
-  <a href="./package.json"><img src="https://img.shields.io/node/v/reasonix.svg?style=flat-square&color=5fa04e&labelColor=161b22&logo=nodedotjs&logoColor=white" alt="node"/></a>
   <a href="https://github.com/esengine/reasonix/stargazers"><img src="https://img.shields.io/github/stars/esengine/reasonix.svg?style=flat-square&color=dbab09&labelColor=161b22&logo=github&logoColor=white" alt="GitHub stars"/></a>
   <a href="https://atomgit.com/esengine/DeepSeek-Reasonix"><img src="https://atomgit.com/esengine/DeepSeek-Reasonix/star/badge.svg" alt="AtomGit stars"/></a>
   <a href="https://github.com/esengine/reasonix/graphs/contributors"><img src="https://img.shields.io/github/contributors/esengine/reasonix.svg?style=flat-square&color=bc8cff&labelColor=161b22&logo=github&logoColor=white" alt="contributors"/></a>
@@ -39,198 +32,237 @@
 </p>
 
 <p align="center">
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/9e931d80-2050-4b10-902e-44970cc133ad.svg" alt="oosmetrics — Agents 速度榜 Top 2"/></a>
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/556d94b3-61b7-486b-baf2-888b9327deab.svg" alt="oosmetrics — LLMs 速度榜 Top 3"/></a>
-  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/0f457d4c-efca-4d15-ad2b-139691ff342c.svg" alt="oosmetrics — CLI 速度榜 Top 3"/></a>
+  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/9e931d80-2050-4b10-902e-44970cc133ad.svg" alt="oosmetrics — Top 2 in Agents by velocity"/></a>
+  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/556d94b3-61b7-486b-baf2-888b9327deab.svg" alt="oosmetrics — Top 3 in LLMs by velocity"/></a>
+  <a href="https://oosmetrics.com/repo/esengine/reasonix"><img src="https://api.oosmetrics.com/api/v1/badge/achievement/0f457d4c-efca-4d15-ad2b-139691ff342c.svg" alt="oosmetrics — Top 3 in CLI by velocity"/></a>
 </p>
 
 <br/>
 
-<h3 align="center">DeepSeek 原生的终端 AI 编程代理。</h3>
-<p align="center">围绕前缀缓存稳定性设计 —— 长会话下 token 成本始终低位运行，可以一直开着。</p>
+<h3 align="center">面向终端的 DeepSeek 原生 AI coding agent。</h3>
+<p align="center">由配置与插件驱动的极薄 harness——单一静态 Go 二进制，围绕 DeepSeek 的前缀缓存调优，长会话也能把 token 成本压低。</p>
 
 <br/>
 
-<p align="center">
-  <img src="docs/assets/hero-terminal.zh-CN.svg" alt="Reasonix code 模式预览 — 助手提出 SEARCH/REPLACE 编辑，未 /apply 不落盘" width="860"/>
-</p>
-
-<br/>
-
-> [!TIP]
-> **缓存稳定不是开关，而是循环要围绕设计的不变量。** 这就是 Reasonix 只支持 DeepSeek 的根本原因 —— 每一层都为 DeepSeek 字节稳定的前缀缓存机制调过。
-
 > [!IMPORTANT]
-> **加入社区 · Community** — 中英双语 Discord，频道包括 `#求助` / `#help`、`#分享` / `#showcase`、`#想法反馈`、贡献者专属 PR 协调区。在群内绑定 GitHub 后自动识别贡献者身份。→ **<https://discord.gg/XF78rEME2D>**
+> **加入社区 · Community** — 双语 Discord，提供安装答疑（`#help` / `#求助`）、工作流展示与功能想法。→ **<https://discord.gg/XF78rEME2D>**
 
-<br/>
+## 特性
+
+- **配置驱动**：provider、agent、启用的工具、插件全部在 `reasonix.toml` 中声明，
+  内核无硬编码模型。
+- **多模型 · 可组合**：DeepSeek（flash/pro）与 MiMo 作为预设内置；任何 OpenAI 兼容
+  端点都只是一条配置。可选让两个模型协同（执行器 + 规划器），各自独立、缓存稳定的 session。
+- **插件驱动**：外部工具以子进程形式运行，通过 stdio JSON-RPC 通信（MCP 兼容）；
+  内置工具在编译期自注册。
+- **零摩擦分发**：`CGO_ENABLED=0` 单二进制；一条命令交叉编译到六个目标平台。
+  唯一依赖是一个 TOML 解析库。
 
 ## 安装
 
-~~~bash
-cd my-project
-npx reasonix code   # 首次运行粘贴 DeepSeek API Key，之后会记住
-~~~
+```sh
+npm i -g reasonix                  # 任意系统;自动拉取对应平台的原生二进制
+brew install esengine/reasonix/reasonix   # macOS
+```
 
-要求 Node ≥ 22。在 macOS · Linux · Windows（PowerShell · Git Bash · Windows Terminal）都跑得顺。[去拿 DeepSeek API Key →](https://platform.deepseek.com/api_keys) · 完整 flag 看 `reasonix code --help`。
+预编译归档(`darwin|linux|windows × amd64|arm64`)和 `SHA256SUMS` 见每个
+[GitHub release](https://github.com/esengine/DeepSeek-Reasonix/releases)。
 
-`npx` 是推荐路径 —— 不用全局安装，每次都拿最新版。如果你天天用、想把 `reasonix` 装到 `PATH`，跑一次 `reasonix update`。
+### 从源码构建
 
-| 命令 | 何时用 |
-|---|---|
-| `reasonix code [dir]` | 编码 agent。**先用这个。** |
-| `reasonix chat` | 纯聊天 —— 不挂文件系统 / shell 工具。 |
-| `reasonix run "task"` | 一次性，结果流到 stdout。适合 shell 管道。 |
-| `reasonix doctor` | 体检：Node 版本、API Key、MCP 接线。 |
-| `reasonix update` | 升级 Reasonix 本身。 |
+```sh
+make build      # -> bin/reasonix(.exe)
+make cross      # -> dist/（darwin|linux|windows × amd64|arm64）
+```
 
-其他子命令（`replay` · `diff` · `events` · `stats` · `index` · `mcp` · `prune-sessions`）在 `reasonix --help` 和 [CLI 参考](https://esengine.github.io/DeepSeek-Reasonix/#cli)。
+## 快速开始
 
-### QQ 通道
-
-Reasonix 可以把现有的 `chat`、`code` 或桌面端会话延伸到 QQ 上，作为远程通道使用；它扩展的是当前会话，不是独立的新运行模式。
-
-- CLI：先启动会话，再执行 `/qq connect`
-- 桌面端：打开 `设置 -> 通用 -> QQ通道`
-
-连接成功后，QQ 消息可以进入当前会话，助手回复会回到 QQ，后续确认和跟进交互也可以继续在 QQ 上完成。
-
-完整配置、桌面端快速上手与排障说明见：[QQ 连接指南](./docs/qq-connect.zh-CN.md)。
-
-<details>
-<summary><strong>切换工作区 · chat vs. code · 写第一个 Skill</strong></summary>
-
-**切换工作区。** Reasonix 把文件系统工具作用域绑定在启动目录，传 `--dir` 可以指别处。中途切换是有意不支持的（消息日志和 memory 路径会和旧根目录混在一起）—— 退出再启动。
-
-~~~bash
-npx reasonix code --dir /path/to/project
-~~~
-
-**`chat` 还是 `code`？** `code` 是默认入口、唯一带文件系统 / shell 工具和 SEARCH/REPLACE 审阅的模式。`chat` 是更轻量的纯对话壳——想要一个挂着 MCP 但没有磁盘权限的“思路助手”时用它。
-
-| 你拿到什么 | `code` | `chat` |
-|---|---|---|
-| 文件系统工具 + `edit_file` | ✓ | — |
-| SEARCH/REPLACE → `/apply` 审阅 | ✓ | — |
-| Shell 工具（带 gate） | ✓ | — |
-| Plan 模式 · `/todo` · `/skill new` · `/mcp add` | ✓ | — |
-| Memory（`remember` / `recall_memory`） | 项目 + 全局 | 仅全局 |
-| 配置里的 MCP · web 搜索 · `ask_choice` | ✓ | ✓ |
-| 编码导向系统提示词 | ✓ | 通用 |
-| Session 作用域 | 按目录 | 共享默认 |
-
-**写第一个 Skill。** 暂无在线市场——自己写。编辑文件（`description:` frontmatter + 正文），然后 `/skill list` 就能看到。frontmatter 加 `runAs: subagent` 会以隔离 subagent 跑，而不是把正文内联进父 prompt。
-
-~~~bash
-/skill new my-skill              # <project>/.reasonix/skills/my-skill.md
-/skill new my-skill --global     # ~/.reasonix/skills，跨项目共用
-~~~
-
-</details>
-
-<br/>
+```sh
+reasonix setup                      # 配置向导 → ./reasonix.toml
+export DEEPSEEK_API_KEY=sk-...  # 或写入 .env（见 .env.example）
+reasonix chat                       # 然后在会话里运行 /init 生成 AGENTS.md（项目记忆）
+reasonix run "把 main.go 里的 TODO 实现掉"
+reasonix run --model mimo-pro "给这个函数补单元测试"
+echo "解释这段代码" | reasonix run
+```
 
 ## 配置
 
-一个全局 JSON 文件 `~/.reasonix/config.json`，加上项目级 `<project>/.reasonix/` 下的覆盖。完整的双语参考 —— 每一个 key、每一条斜杠命令、skills / memory / hooks 在磁盘上的形状 —— 都在这里：
+优先级：**flag > `./reasonix.toml` > `~/.config/reasonix/config.toml` > 内置默认值**。
+密钥经环境变量通过 `api_key_env` 注入，绝不写入配置文件。
 
-> 📘 **[配置指南](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh)** · [English](https://esengine.github.io/DeepSeek-Reasonix/configuration.html)
+```toml
+default_model = "deepseek-flash"   # 执行器；设 [agent].planner_model 可加规划器
+# language    = "zh"               # 界面语言；为空则按 $LANG / $REASONIX_LANG 自动检测
 
-| 主题 | 速读 |
-|---|---|
-| [MCP 服务器](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#mcp) | stdio · SSE · Streamable HTTP。`config.json` 和 `--mcp` 共用同一种 spec 格式。 |
-| [Skills](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#skills) | 模型可以调用的 markdown 剧本。`inline` 或 `subagent` 两种模式。 |
-| [Memory](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#memory) | 用户私有的知识，钉进前缀。`user` / `feedback` / `project` / `reference` 四类。 |
-| [Hooks](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#hooks) | 生命周期事件触发的 shell 命令。`PreToolUse`（拦截）· `PostToolUse` · `UserPromptSubmit` · `Stop`。 |
-| [权限](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#permissions) | 按工作区的 shell 白名单，精确前缀匹配。 |
-| [Web 搜索](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#search) | 默认 Bing；可用 `/search-engine` 切到百度 AI Search、自托管 SearXNG、Metaso、Tavily、Perplexity、Exa、Brave 或 Ollama。 |
-| [语义索引](https://esengine.github.io/DeepSeek-Reasonix/configuration.html?lang=zh#index) | `reasonix index` —— 本地 Ollama，或任何 OpenAI 兼容的 embedding 接口。 |
+[agent]
+# planner_model = "mimo-pro"          # 可选的低频规划器
+# subagent_model = "deepseek-pro"     # runAs=subagent skill 的默认模型
+# subagent_models = { review = "deepseek-pro", security_review = "deepseek-pro" }
+auto_plan = "off"                  # off|on；off 表示计划模式仅手动开启
+# auto_plan_classifier = "deepseek-flash"   # 可选；只在边界任务上调用
 
-<br/>
+[[providers]]
+name        = "deepseek-flash"
+kind        = "openai"
+base_url    = "https://api.deepseek.com"
+model       = "deepseek-v4-flash"
+api_key_env = "DEEPSEEK_API_KEY"
+# 还有预设：deepseek-pro、mimo-pro（mimo-v2.5-pro）、mimo-flash（mimo-v2-flash） @ api.xiaomimimo.com/v1
 
-## Reasonix 的不同之处
+[tools]
+enabled = []   # 省略/为空 = 全部内置工具
 
-整个循环围绕三根支柱组织。每一根解决的都是通用 agent 框架根本看不见的问题 —— 因为它们是为另一种缓存机制设计的。
+[skills]
+# paths = ["~/my-skills", "../shared/skills"]   # 额外的自定义技能目录
+# disabled_skills = ["review"]                  # 隐藏技能，直到 /skill enable <name>
 
-<sub align="center">
+[permissions]
+mode  = "ask"                                # 无规则命中时 writer 的兜底：ask|allow|deny
+deny  = ["bash(rm -rf*)", "bash(git push*)"] # 任何模式下都硬阻断
+allow = ["bash(go test*)"]                   # 从不询问
 
-各支柱完整说明 → [Pillar 1 — 缓存优先循环](./docs/ARCHITECTURE.md#pillar-1--cache-first-loop) · [Pillar 2 — 工具调用修复](./docs/ARCHITECTURE.md#pillar-2--tool-call-repair) · [Pillar 3 — 成本控制](./docs/ARCHITECTURE.md#pillar-3--cost-control-v06)
+[sandbox]
+# workspace_root = ""          # 文件写工具被限制在此目录；留空 = 当前目录
+# allow_write    = ["/tmp"]    # write_file/edit_file/multi_edit 额外可写的目录
 
-</sub>
+[[plugins]]
+name    = "example"
+command = "reasonix-plugin-example"
+```
 
-<br/>
+权限逐次调用把关：`deny` > `ask` > `allow` > 兜底（只读工具永远 allow，writer 落到
+`mode`）。`reasonix chat` 会在 writer 调用前征求同意（`1` 本次 · `2` 本会话 · `3` 拒绝，兼容 `y/a/n`）；
+`reasonix run` 保持自主运行但仍然遵守 `deny`。完整 schema 与契约见
+[`docs/SPEC.md`](docs/SPEC.md)。
 
-## 能力一览
+权限是**策略**（哪些调用放行/询问），**沙盒**是**强制**：文件写工具
+（`write_file` / `edit_file` / `multi_edit`）拒绝 `[sandbox] workspace_root`
+之外的任何路径（默认当前目录，编辑不出项目），并解析符号链接与 `..`，使链接无法
+打洞越界。读不受限。`bash` 本身在 macOS 默认进沙盒（`[sandbox] bash`，Seatbelt）：
+命令只能写这些 root（外加临时目录与工具链缓存），`[sandbox] network` 为真时才能联网；
+其它平台暂回退为不沙盒运行（越界问一次与 Linux 支持见 `docs/SPEC.md` §9）。
 
-<p align="center">
-  <img src="docs/assets/feature-grid.zh-CN.svg" alt="Reasonix 能力一览 — cell-diff 渲染器、MCP、计划模式、权限、仪表盘、持久化会话、Hooks/Skills/Memory、语义检索、自动 checkpoint、/effort 旋钮、transcript 重放、事件日志" width="880"/>
-</p>
+### 插件（MCP）
 
-<br/>
+Reasonix 是一个 MCP 客户端。`[[plugins]]` 的 `type` 选择传输：`stdio`（默认）启动本地子进
+程（`command`/`args`/`env`）；`http`（Streamable HTTP）连接远程 `url`，可带静态
+`headers`（`${VAR}` / `${VAR:-default}` 从环境展开，密钥不入文件）。工具以
+`mcp__<server>__<tool>` 暴露给模型，与 Claude Code 一致；声明 MCP `readOnlyHint: true`
+的工具会参与并行调度并命中权限层的只读默认放行。
 
-## 横向对比
+服务器的 **prompts** 会暴露成 `/mcp__<server>__<prompt>` 斜杠命令（命令后空格分隔参
+数）；**resources** 通过在消息里写 `@<server>:<uri>` 拉入；`/mcp` 列出已连接服务器及
+各自暴露的内容。`make build` 还会产出 `bin/reasonix-plugin-example`——一个可直接运行的
+stdio 参考实现（`echo`、`wordcount`、一个 `review` prompt、一个 style-guide 资源），
+可照抄。
 
-|                            | Reasonix          | Claude Code       | Cursor              | Aider              |
-|----------------------------|-------------------|-------------------|---------------------|--------------------|
-| 后端                       | DeepSeek          | Anthropic         | OpenAI / Anthropic  | 任意（OpenRouter） |
-| 协议                       | **MIT**           | 闭源              | 闭源                | Apache 2           |
-| 单任务成本                 | **低**            | 高                | 订阅 + 用量         | 不一               |
-| DeepSeek 前缀缓存          | **专门工程化**    | 不适用            | 不适用              | 偶发命中           |
-| 内嵌 web 仪表盘            | 支持              | —                 | 不适用 (IDE)        | —                  |
-| 持久化的工作区会话         | 支持              | 部分              | 不适用              | —                  |
-| 计划模式 · MCP · Hooks     | 支持              | 支持              | 支持                | 部分               |
-| 开放社区共建               | 支持              | —                 | —                   | 支持               |
+```toml
+[[plugins]]                       # 本地 stdio 服务器
+name    = "example"
+command = "reasonix-plugin-example"
 
-实测缓存命中率、成本、方法论看 [`benchmarks/`](./benchmarks/) —— 这些数会随模型定价变化，所以归在 harness 里，不进 README。
+[[plugins]]                       # 远程 Streamable HTTP 服务器
+name    = "stripe"
+type    = "http"
+url     = "https://mcp.stripe.com"
+headers = { Authorization = "Bearer ${STRIPE_KEY}" }
+```
 
-<br/>
+**已有 Claude Code 的 `.mcp.json`？** 直接放到项目根目录，Reasonix 会原样读取——其
+`mcpServers` 规范（`command`/`args`/`env`、`type`/`url`/`headers`、`${VAR}` 展开）
+与 `[[plugins]]` 字段一一对应。两处来源会合并加载；同名时以 `reasonix.toml` 为准。
 
-## 文档
+```json
+{
+  "mcpServers": {
+    "filesystem": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path"] },
+    "stripe": { "type": "http", "url": "https://mcp.stripe.com", "headers": { "Authorization": "Bearer ${STRIPE_KEY}" } }
+  }
+}
+```
 
-- [**架构**](./docs/ARCHITECTURE.md) —— 四大支柱、缓存优先循环、思维提取、脚手架
-- [**CLI 参考**](./docs/CLI-REFERENCE.md) —— 每个 shell 子命令、每个 slash 命令、每个快捷键
-- [**QQ 连接指南**](./docs/qq-connect.zh-CN.md) —— CLI 首次连接流程、桌面端入口和 QQ 开放平台凭据
-- [**基准测试**](./benchmarks/) —— τ-bench-lite harness、transcript、成本方法论
-- [**官方网站**](https://esengine.github.io/DeepSeek-Reasonix/) —— 入门、Dashboard 设计稿、TUI 设计稿
-- [**贡献指南**](./CONTRIBUTING.md) —— 注释规则、错误处理、用现成库不手写
-- [**行为准则**](./CODE_OF_CONDUCT.md) · [**安全策略**](./SECURITY.md)
+**从 `0.x` 升级？** 旧的 `~/.reasonix/config.json` 仍会被读取(读其 `mcpServers`、并遵从
+`mcpDisabled`),作为最低优先级来源——所以 MCP 服务器照常可用;方便时再把它们挪进
+`reasonix.toml` 的 `[[plugins]]` 或 `.mcp.json`。
 
-<br/>
+### 斜杠命令
 
-## 社区
+`reasonix chat` 里,内置命令(`/compact`、`/new`、`/rewind`、`/tree`、`/branch`、`/switch`、`/todo`、`/model`、`/effort`、`/mcp`、`/help`)在本地执行。
+`/tree` 查看已保存的对话分支,`/branch [name]` 从当前对话末端分支,`/branch <turn> [name]`
+从较早的 checkpoint 轮次分支,`/switch <id|name>` 切换到另一个分支。**自定义命令**
+是放在 `.reasonix/commands/`(项目)或 `~/.config/reasonix/commands/`(用户)下的 Markdown 文件——
+`review.md` 即 `/review`,子目录构成命名空间(`git/commit.md` → `/git:commit`)。文件正文
+是 prompt 模板,调用即作为一轮对话发出。
 
-> [!NOTE]
-> Reasonix 是开源、社区共建的项目。文末"致谢"那面贡献者墙上的每一个头像，都对应一次真实合并的 PR。
+```markdown
+---
+description: Review the staged diff
+argument-hint: [focus-area]
+---
+Review the staged diff. Focus on $ARGUMENTS, list bugs with file:line.
+```
 
-给新手准备的入门 issue —— 每个都带背景说明、代码定位、验收标准、提示 —— 全部挂在 [`good first issue`](https://github.com/esengine/reasonix/labels/good%20first%20issue) 标签下。挑任意一个还没人认领的就行。
+`$ARGUMENTS` 展开为全部空格分隔参数,`$1`…`$N` 为位置参数。MCP prompts 也以
+`/mcp__<server>__<prompt>` 形式出现在这里。
 
-**正在征集意见的 Discussions：**
+### @ 引用
 
-- [#20 · CLI / TUI 设计](https://github.com/esengine/reasonix/discussions/20) —— 哪里坏了、哪里少东西、哪里你会怎么改？
-- [#21 · Dashboard 设计](https://github.com/esengine/reasonix/discussions/21) —— 对着[设计稿](https://esengine.github.io/DeepSeek-Reasonix/design/agent-dashboard.html)拍砖
-- [#22 · 未来功能愿望单](https://github.com/esengine/reasonix/discussions/22) —— 你希望 Reasonix 长出什么功能？
+在消息里写 `@` 引用,Reasonix 会在发送前解析成带标签的上下文块:`@path/to/file`(或
+`@dir`)注入本地文件内容(或目录清单),`@<server>:<uri>` 注入 MCP 资源。本地路径**只有
+真实存在**时才当作引用,普通 `@mention` 保持原文。敲 `/` 或 `@` 会弹出补全菜单——斜杠
+命令,或**逐层**的文件导航(一次只列当前一层目录、可下钻进子目录)外加 MCP 资源。
 
-**正在使用 Reasonix，愿意让更多人了解它？** 欢迎将相关博客、文章、截图、演讲或视频发布到 [**Show and tell**](https://github.com/esengine/reasonix/discussions/categories/show-and-tell)。项目没有营销预算，新用户主要通过社区口碑找到这里。持续参与传播的用户将获得下方这枚徽章，颁发后会展示在贡献者墙旁：
+### 双模型协同（可选）
 
-<p align="center">
-  <a href="https://github.com/esengine/reasonix/discussions/categories/show-and-tell">
-    <img src="https://img.shields.io/badge/REASONIX-📣%20ADVOCATE-c4b5fd?style=for-the-badge&labelColor=0d1117" alt="Reasonix Advocate 徽章 —— 授予持续参与传播的用户"/>
-  </a>
-</p>
+`reasonix setup` 刻意保持首次体验极简：选 provider → 输入 key（所选 provider 的所有
+SKU 都会启用）。若要让两个模型协同（执行器 + 规划器，各自独立、缓存稳定的
+session），向导后手动在 `reasonix.toml` 加一行即可：
 
-**第一次提 PR 之前**：先读 [`CONTRIBUTING.md`](./CONTRIBUTING.md) —— 短小、严格的项目规则（注释、错误处理、用现成库不手写）。`tests/comment-policy.test.ts` 静态强制执行注释那部分，`npm run verify` 是 push 前的闸。参与本项目即同意 [行为准则](./CODE_OF_CONDUCT.md)。安全相关问题请走 [SECURITY.md](./SECURITY.md)。
+```toml
+[agent]
+planner_model = "deepseek-pro"   # 作为低频规划器
+```
 
-<br/>
+Subagent skills 默认继承执行器模型。设置 `subagent_model` 可让它们统一走另一个已配置
+模型；设置 `subagent_models` 则只覆盖 `review`、`security_review` 等指定 skill。
 
-## 不做的事
+交互式前端中，计划模式默认手动开启。设置 `agent.auto_plan = "on"` 后，看起来复杂
+的任务会自动进入 plan mode：Reasonix 先只读生成计划，待用户批准后才
+编辑文件或执行有副作用的命令。`auto_plan_classifier` 可以指定便宜的 provider，例如
+`deepseek-flash`；它只在边界输入上调用，分类失败会回退到启发式规则。也可以用
+`reasonix chat` 里的 `/auto-plan off|on` 修改用户级设置，或在 shell/脚本里用
+`reasonix config auto-plan off|on`。只有明确想写项目级覆盖时，才给 shell 命令加
+`--local`。
 
-> [!IMPORTANT]
-> Reasonix 是有立场的。有些事它故意 *不做* —— 列在这里方便你为自己的工作挑对工具。
+## 架构
 
-- **多供应商灵活性。** 故意只做 DeepSeek。绑死一个后端是 feature，不是限制。
-- **IDE 集成。** 终端优先。diff 在 `git diff`，文件树在 `ls`。仪表盘是 TUI 的伴生，不是 Cursor 的替代。
-- **追最难的 reasoning 榜单。** Claude Opus 在某些榜单上还是赢家。DeepSeek 在编程任务上有竞争力；如果你的工作是"解一个 PhD 级证明"而不是"修个 auth bug"，先用 Claude。
-- **完全离线 / 永远免费。** Reasonix 需要付费的 DeepSeek API Key。要离线 / 零成本，看 Aider + Ollama 或 [Continue](https://continue.dev)。
+三层可扩展性，全部藏在内核按名解析的 registry 之后：
+
+1. **Registry**：`Provider` 与 `Tool` 是接口；内核没有 `switch model`。
+2. **编译期内置**：provider（`provider/openai`）和 tool（`tool/builtin`）通过
+   `init()` 自注册，`main` 用 blank import 拉入。新增内置 = 一个文件 + 一行 import。
+3. **运行时插件**：配置里声明的可执行文件，通过 stdin/stdout 上的
+   newline-delimited JSON-RPC 2.0（MCP stdio 约定）通信，每个远程 tool 适配成
+   `Tool` 接口。
+
+## 状态
+
+已完成：基于 registry 的 provider/tool、OpenAI 兼容流式 + 工具调用（429/5xx 有界重
+试）、九个内置工具（read_file、write_file、edit_file、multi_edit、bash、ls、glob、
+grep、web_fetch）、TOML 配置、交互式 `reasonix setup` 向导、双模型协同（执行器 + 规划器，
+各自独立、缓存稳定的 session）、低频上下文压缩、子 agent（`task`）、bubbletea 聊天
+TUI（markdown、plan mode、上下文仪表盘、`/compact` `/new` `/tree` `/branch` `/switch`）、会话持久化 + 恢复、
+逐次调用**权限**（allow/ask/deny 规则；chat 在 writer 前询问，deny 在各模式硬阻断）、
+**工作区沙盒**（把文件写工具限制在项目内，符号链接/`..` 安全）、
+MCP 客户端——**stdio + Streamable HTTP** 传输、工具（`mcp__server__tool`,支持
+`readOnlyHint`）、prompts（斜杠命令）、resources（`@` 引用）、`/mcp`，可经
+`[[plugins]]` 或 Claude 风格的项目 `.mcp.json` 配置——自定义斜杠命令
+（`.reasonix/commands/*.md`）、`@file` / `@resource` 引用、外加可运行的参考插件
+（`cmd/reasonix-plugin-example`）、harness 主循环、CLI。chat 在终端普通缓冲区运行(原生
+scrollback)并带 `/` 与 `@` 输入补全。后续:给 `bash` 套 OS 级沙盒（macOS Seatbelt /
+Linux bubblewrap，"盒子里放行、边界上询问"）、Anthropic 原生 provider、MCP OAuth +
+legacy SSE。见 `docs/SPEC.md` §9。
 
 <br/>
 
@@ -275,8 +307,8 @@ npx reasonix code --dir /path/to/project
 - [**wade19990814-hue**](https://github.com/wade19990814-hue)
 - [**wviana**](https://github.com/wviana)（Wesley Viana）
 
-另外特别感谢 [**Bernardxu123**](https://github.com/Bernardxu123) 设计的项目 logo
-（见 [`docs/brand/`](./docs/brand/)），以及 [AIGC Link](https://xhslink.com/m/80ngts127cA) 在小红书上的推广。
+另外特别感谢 [**Bernardxu123**](https://github.com/Bernardxu123) 设计的项目 logo，
+以及 [AIGC Link](https://xhslink.com/m/80ngts127cA) 在小红书上的推广。
 
 <p align="center">
   <a href="https://github.com/esengine/DeepSeek-Reasonix/graphs/contributors">

REASONIX.md

@@ -1,73 +1,28 @@
-# Reasonix — working knowledge
+# Reasonix project memory
 
-TypeScript project. DeepSeek-native coding agent, cache-first loop.
-MIT-licensed. Node ≥22 required.
-
-## Stack
-
-- **Language** — TS 5.6+, ES2022, ESM (`"type": "module"`)
-- **CLI** — Commander.js + Ink 5 (React 18) TUI
-- **Test** — Vitest 2.x
-- **Lint / Format** — Biome 1.9 (2-space, double quotes, semicolons always, 100 width)
-- **Build** — tsup (bundle), `tsx` (dev runner)
-- **MCP** — stdio + SSE transports, in-process fake in tests
-
-## Layout
-
-| Path | What |
-|---|---|
-| `src/cli/` | CLI entry + commands (`chat.tsx`, `code.tsx`, `diff.ts`, etc.) + Ink TUI in `ui/` |
-| `src/tools/` | Tool defs (filesystem, shell, MCP, plan, subagent, web, workspace) |
-| `src/mcp/` | MCP client, transports (stdio, SSE), registry, spec |
-| `src/repair/` | Tool-call repair pipeline (flatten, scavenge, storm, truncation) |
-| `src/index/` | Semantic vector index |
-| `src/code/` | SEARCH/REPLACE edit-block parser + apply gate |
-| `src/core/` | Event-log kernel — `events.ts` (Event union), `reducers.ts` (pure projections), `eventize.ts` |
-| `src/ports/` | Port interfaces — ModelClient, ToolHost, EventSink, MemoryStore, HookRunner, CheckpointStore |
-| `src/adapters/` | Concrete adapters for the ports (e.g. `event-sink-jsonl.ts`, `event-source-jsonl.ts`) |
-| `src/frame/` | Frame compiler (cell grid → ANSI) used by the TUI log renderer |
-| `src/memory/` | Project / session / user / runtime memory stores |
-| `src/transcript/` | Transcript log (write), diff, replay |
-| `src/telemetry/` | Usage records + cross-session stats |
-| `src/server/` | Dashboard HTTP server + REST API |
-| `tests/` | Vitest tests, flat `*.test.ts` |
-| `examples/` | `basic-chat.ts`, `mcp-server-demo.ts`, etc. |
-| `benchmarks/` | Harvest + tau-bench harnesses |
-| `dashboard/` | Compiled dashboard SPA assets |
-| `data/` | Tokenizer data (`deepseek-tokenizer.json.gz`) |
-| `dist/` | Build output — **do not edit** |
-| `.github/` | CI + issue / PR templates |
-
-## Commands
-
-```sh
-npm run build       # tsup → dist/
-npm run dev         # tsx src/cli/index.ts
-npm run chat        # tsx src/cli/index.ts chat
-npm run test        # vitest run
-npm run test:watch  # vitest
-npm run lint        # biome check src tests
-npm run lint:fix    # biome check --write src tests
-npm run format      # biome format --write src tests
-npm run typecheck   # tsc --noEmit
-```
-
-`prepublishOnly`: lint → typecheck → test → build.
+This file is loaded into every session's system prompt (the cache-stable prefix),
+so keep it concise and durable — it is the project's standing instructions to the
+agent. It is the Reasonix analog of Claude Code's CLAUDE.md.
 
 ## Conventions
 
-- **Imports** — explicit `import type` for type-only imports (Biome `useImportType: warn`). Direct relative imports within project, no barrel re-exports.
-- **Exports** — named exports only; no `export default`. Entry: `src/index.ts`.
-- **Tests** — vitest `describe`/`it`/`expect`, no globals. Naming: `<module>.test.ts` flat in `tests/`.
-- **JSX** — `.tsx` for Ink components. `jsx: "react"` in tsconfig.
-- **TypeScript** — `strict`, `noUncheckedIndexedAccess`, `noImplicitOverride`. Tools accept `ToolCallContext` (abort signal).
-- **MCP** — All transports implement `McpTransport` interface. Tools registered via registry at startup.
-- **Changelog** — Keep a Changelog format. Semver.
+- Go kernel under `internal/`; each package owns one concern and documents it in a
+  package comment. Match the surrounding comment density and idiom when editing.
+- One transport-agnostic `control.Controller` sits behind every frontend (chat
+  TUI, HTTP/SSE serve, Wails desktop). Add behavior to the controller, not a
+  frontend, so all three inherit it.
+- Cache-first: the system-prompt prefix (base prompt + tools + memory) must stay
+  byte-stable across turns so DeepSeek's automatic prefix cache stays warm. Never
+  mutate it mid-session — ride the turn tail instead (see `control.Compose`).
 
-## Watch out for
+## Memory
 
-- **This IS Reasonix** — edits to `src/loop.ts`, `src/repair/`, `src/tools/`, `src/mcp/` affect every session. Test before publishing.
-- **SEARCH must match byte-for-byte** — the edit-gate in `src/code/edit-blocks.ts` enforces exact match. Trailing whitespace or wrong indent = mismatch.
-- **`dist/`** is generated by `tsup`. Never hand-edit.
-- **`.reasonix/semantic/`** is auto-generated vector index. Never hand-edit.
-- **`sessions/` and `.reasonix/sessions/`** are user-private, git-ignored (per `.gitignore`).
+- Hierarchical docs: `REASONIX.md` (this file, committed/shared), `REASONIX.local.md`
+  (personal, git-ignored), user-global `~/.config/reasonix/REASONIX.md`, and any
+  `REASONIX.md` in an ancestor dir. `AGENTS.md` is accepted as a fallback name.
+- `@path` on its own line imports another file's contents.
+- `#<note>` in chat quick-adds a line here. The `remember` tool saves durable
+  facts to the per-project auto-memory store (frontmatter files + `MEMORY.md`
+  index), which loads into the prefix on the next session.
+
+## Notes

SECURITY.md

@@ -1,40 +0,0 @@
-# Security Policy
-
-If you find a security issue in Reasonix, please report it privately rather than opening a public issue or discussion thread.
-
-## How to report
-
-Email <359807859@qq.com> with:
-
-- a clear description of the issue
-- steps that reproduce it (a minimal repro is fine)
-- the version (`reasonix --version`) and platform you observed it on
-
-You'll get an acknowledgement within a few days, and a fix or mitigation as soon as the maintainer can land it. If you'd like attribution in the release notes when the fix ships, say so in your report — the default is a quiet patch.
-
-## Supported versions
-
-Only the latest published minor of `reasonix` on npm is actively maintained. If you're on something older, please reproduce on the latest before reporting.
-
-## Scope
-
-**In scope:**
-
-- The published `reasonix` npm package and its CLI / TUI
-- The dashboard SPA shipped under `dashboard/` and the local HTTP server that serves it
-- The shell sandbox, edit gate, and tool dispatcher in `src/`
-
-**Out of scope:**
-
-- Third-party MCP servers attached via `--mcp` (report to those projects)
-- Misconfiguration of the user's own DeepSeek API key, environment, or shell profile
-- Vulnerabilities in upstream Node.js or in the DeepSeek API itself
-- Denial-of-service via deliberately oversized prompts or tool inputs (Reasonix is a single-user CLI; there's no multi-tenant boundary to defend)
-
-## Hardening notes
-
-A few practical reminders for users running Reasonix:
-
-- API keys live in `~/.reasonix/config.json`. Treat that file like any other credential store.
-- `run_command` and the `!` shell shortcut respect a permission allowlist; the safe default is `ask` on anything not pre-approved. Don't set `editMode: yolo` on machines that hold secrets you'd regret leaking.
-- Hooks (`PreToolUse`, etc.) execute arbitrary shell scripts the user has configured. Audit `.reasonix/settings.json` before running Reasonix in a directory you didn't author.

benchmarks/README.md

@@ -1,115 +0,0 @@
-# Benchmarks
-
-This is where validation lives. The v0.1 milestone gates on a reproducible
-tool-use eval that compares, on the same tasks:
-
-1. **Baseline** — a deliberately cache-hostile agent (fresh timestamp +
-   shuffled tool spec each turn), representative of how generic frameworks
-   wire up DeepSeek.
-2. **Reasonix** — the same tools and system prompt, driven through
-   `CacheFirstLoop` so the byte prefix stays stable turn-over-turn.
-
-Both modes share the same `DeepSeekClient`, so the *only* meaningful
-difference is prefix stability — any cache-hit / cost gap is attributable to
-Pillar 1 of the architecture, nothing else.
-
-## Scope — this is τ-bench-*lite*
-
-We don't ship a full port of [Sierra's τ-bench](https://github.com/sierra-research/tau-bench)
-(airline + retail, Python). Instead:
-
-- `tau-bench/tasks.ts` hand-authors 8 retail-flavored multi-turn tasks
-  that exercise tool use, identity verification, refusal, and mid-conversation
-  goal change.
-- The task schema (`tau-bench/types.ts`) mirrors τ-bench's shape — stateful
-  tools, an LLM user simulator, end-state DB predicates — so real upstream
-  tasks can later drop in without harness changes.
-- All success predicates are **deterministic DB checks**, not LLM judges.
-  Refusal tasks pass iff the DB is unchanged.
-
-## Files
-
-```
-tau-bench/
-├── types.ts       — TaskDefinition / RunResult / BenchReport shapes
-├── db.ts          — tiny in-memory WorldState + cloneDb
-├── tasks.ts       — the 8 seed tasks + shared tool factories
-├── user-sim.ts    — LLM user simulator (V3, T=0.1)
-├── baseline.ts    — naive cache-hostile agent runner
-├── runner.ts      — orchestrates user-sim × agent × task × mode
-└── report.ts      — turns a results-*.json into a report.md
-```
-
-## Quickstart
-
-```bash
-# dry-run: no API calls, just validate the harness is wired up
-npx tsx benchmarks/tau-bench/runner.ts --dry
-
-# full run: both modes, all tasks, 1 repeat
-export DEEPSEEK_API_KEY=sk-...
-npx tsx benchmarks/tau-bench/runner.ts
-
-# tighten variance: 3 repeats per task
-npx tsx benchmarks/tau-bench/runner.ts --repeats 3
-
-# narrow to one task while iterating
-npx tsx benchmarks/tau-bench/runner.ts --task t01_address_happy --verbose
-
-# render the report
-npx tsx benchmarks/tau-bench/report.ts benchmarks/tau-bench/results-<date>.json
-
-# emit per-run transcripts so you can reasonix replay / diff them
-npx tsx benchmarks/tau-bench/runner.ts --transcripts-dir ./transcripts
-npx reasonix diff \
-  ./transcripts/t01_address_happy.baseline.r1.jsonl \
-  ./transcripts/t01_address_happy.reasonix.r1.jsonl \
-  --md diff.md
-```
-
-The runner writes `benchmarks/tau-bench/results-<iso-timestamp>.json`. Point
-`report.ts` at it (or pass `--out report.md` to override the output path).
-
-When `--transcripts-dir <path>` is set, each `(task, mode, repeat)` run also
-writes a `<taskId>.<mode>.r<n>.jsonl` transcript into that directory —
-these carry per-turn `usage`, `cost`, and (for Reasonix) the
-`prefixHash`, so `reasonix replay` and `reasonix diff` can rebuild the
-economics offline.
-
-## CLI flags
-
-| flag | default | meaning |
-|---|---|---|
-| `--task <id>` | all | run only one task by id |
-| `--mode baseline` \| `reasonix` | both | restrict to one mode |
-| `--repeats <N>` | 1 | repeat each (task, mode) pair N times |
-| `--model <id>` | deepseek-chat | agent model |
-| `--user-model <id>` | deepseek-chat | user-simulator model |
-| `--out <path>` | `results-<ts>.json` | results file path |
-| `--transcripts-dir <path>` | off | write one transcript per run for replay/diff |
-| `--dry` | off | skip the LLM; only wire-check |
-| `--verbose` \| `-v` | off | print every user / agent / tool line |
-
-## What a run costs
-
-A full run (8 tasks × 2 modes × 1 repeat) does on the order of 30–60
-DeepSeek V3 calls — well under $0.05 at current pricing. `--repeats 3`
-triples that.
-
-## Adding tasks
-
-1. Add a `TaskDefinition` to `tau-bench/tasks.ts`. Reuse the tool factories
-   defined at the top of that file, or add new ones (remember: factories so
-   tools close over the *per-run* db snapshot).
-2. Make the `check` predicate check the end-state DB, not the agent's text —
-   agents phrase things differently on every run.
-3. Run `--task <your_id> --verbose` to eyeball the transcript.
-
-Non-goals (for this harness):
-
-- LLM-as-judge — brittle and expensive, DB predicates are enough.
-- Streaming comparison — the harness uses `stream: false` in Reasonix mode
-  so both runners make the exact same request shape.
-- Claude head-to-head — we estimate Claude's cost from token counts using
-  Sonnet 4.6 pricing (see `src/telemetry.ts`); running Claude for real is
-  out of scope.

benchmarks/behavior-stability/harness.ts

@@ -1,39 +0,0 @@
-/** Evaluation harness — runs scenarios and collects results. */
-
-import type { EvalReport, EvalResult, EvalScenario } from "./types.js";
-
-export async function runScenarios(
-  scenarios: EvalScenario[],
-  opts: {
-    onResult?: (r: EvalResult) => void;
-  } = {},
-): Promise<EvalReport> {
-  const results: EvalResult[] = [];
-  for (const sc of scenarios) {
-    const t0 = Date.now();
-    let result: EvalResult;
-    try {
-      result = await sc.run();
-      result.durationMs = Date.now() - t0;
-    } catch (err) {
-      result = {
-        scenarioId: sc.id,
-        pass: false,
-        durationMs: Date.now() - t0,
-        details: `UNEXPECTED ERROR: ${(err as Error).message}`,
-      };
-    }
-    results.push(result);
-    opts.onResult?.(result);
-  }
-
-  // Dynamic import avoids bundling VERSION into the harness module.
-  const { VERSION } = await import("../../src/version.js");
-  return {
-    meta: {
-      date: new Date().toISOString(),
-      reasonixVersion: VERSION,
-    },
-    results,
-  };
-}

benchmarks/behavior-stability/report.ts

@@ -1,81 +0,0 @@
-/** Render behavior-stability results.json → markdown report. */
-
-import { readFileSync, writeFileSync } from "node:fs";
-import { pathToFileURL } from "node:url";
-import type { EvalReport } from "./types.js";
-
-interface CliArgs {
-  input: string;
-  outPath: string;
-}
-
-function parseArgs(argv: string[]): CliArgs {
-  const out: CliArgs = { input: "", outPath: "benchmarks/behavior-stability/report.md" };
-  const positional: string[] = [];
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i];
-    if (a === "--out") out.outPath = argv[++i] ?? out.outPath;
-    else if (a && !a.startsWith("--")) positional.push(a);
-  }
-  out.input = positional[0] ?? "";
-  if (!out.input) {
-    throw new Error(
-      "usage: npx tsx benchmarks/behavior-stability/report.ts <results.json> [--out report.md]",
-    );
-  }
-  return out;
-}
-
-function renderHeader(report: EvalReport): string {
-  return `# Reasonix Behavior-Stability Evaluation
-
-**Date:** ${report.meta.date}
-**Version:** ${report.meta.reasonixVersion}
-`;
-}
-
-function renderSummary(report: EvalReport): string {
-  const passed = report.results.filter((r) => r.pass).length;
-  const total = report.results.length;
-  const pct = total > 0 ? ((passed / total) * 100).toFixed(0) : "0";
-
-  const rows = report.results.map((r) => {
-    const icon = r.pass ? "✅" : "❌";
-    return `| ${icon} | \`${r.scenarioId}\` | ${r.durationMs}ms | ${r.details} |`;
-  });
-
-  return `## Summary
-
-**${passed}/${total} passed (${pct}%)**
-
-| status | scenario | time | details |
-|---|---:|---:|:---|
-${rows.join("\n")}
-`;
-}
-
-function render(report: EvalReport): string {
-  return [renderHeader(report), renderSummary(report)].join("\n\n");
-}
-
-function isMain(): boolean {
-  const entry = process.argv[1];
-  if (!entry) return false;
-  return import.meta.url === pathToFileURL(entry).href;
-}
-
-async function main(): Promise<void> {
-  const args = parseArgs(process.argv.slice(2));
-  const raw = readFileSync(args.input, "utf8");
-  const report = JSON.parse(raw) as EvalReport;
-  const md = render(report);
-  writeFileSync(args.outPath, md, "utf8");
-  console.log(`Wrote ${args.outPath} (${report.results.length} results)`);
-}
-
-if (isMain()) {
-  main().catch((err) => {
-    console.error(err);
-    process.exit(1);
-  });
-}

benchmarks/behavior-stability/runner.ts

@@ -1,87 +0,0 @@
-/** CLI runner for behavior-stability evaluation. */
-
-import { mkdirSync, writeFileSync } from "node:fs";
-import { dirname, join } from "node:path";
-import { pathToFileURL } from "node:url";
-import { loadDotenv } from "../../src/index.js";
-import { ALL_SCENARIOS, API_SCENARIOS, LOCAL_SCENARIOS } from "./scenarios.js";
-import { runScenarios } from "./harness.js";
-import type { EvalScenario } from "./types.js";
-
-loadDotenv();
-
-interface CliArgs {
-  local: boolean;
-  api: boolean;
-  outPath: string | null;
-}
-
-function parseArgs(argv: string[]): CliArgs {
-  const out: CliArgs = { local: false, api: false, outPath: null };
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i];
-    if (a === "--local") out.local = true;
-    else if (a === "--api") out.api = true;
-    else if (a === "--out") out.outPath = argv[++i] ?? null;
-  }
-  // Default to local-only when neither flag is provided.
-  if (!out.local && !out.api) out.local = true;
-  return out;
-}
-
-function pickScenarios(args: CliArgs): EvalScenario[] {
-  const scenarios: EvalScenario[] = [];
-  if (args.local) scenarios.push(...LOCAL_SCENARIOS);
-  if (args.api) scenarios.push(...API_SCENARIOS);
-  return scenarios;
-}
-
-function isMain(): boolean {
-  const entry = process.argv[1];
-  if (!entry) return false;
-  return import.meta.url === pathToFileURL(entry).href;
-}
-
-async function main(): Promise<void> {
-  const args = parseArgs(process.argv.slice(2));
-  const scenarios = pickScenarios(args);
-
-  if (scenarios.length === 0) {
-    console.log("No scenarios selected. Use --local, --api, or both.");
-    process.exit(0);
-  }
-
-  const categories = new Set(scenarios.map((s) => s.category));
-  console.log(`Running ${scenarios.length} scenario(s) [${[...categories].join(", ")}] …\n`);
-
-  const report = await runScenarios(scenarios, {
-    onResult: (r) => {
-      const icon = r.pass ? "✅" : "❌";
-      console.log(`${icon} ${r.scenarioId} (${r.durationMs}ms)`);
-      console.log(`   ${r.details}\n`);
-    },
-  });
-
-  const passed = report.results.filter((r) => r.pass).length;
-  const total = report.results.length;
-  console.log(`────────────────────────────────────────`);
-  console.log(`Results: ${passed}/${total} passed`);
-  console.log(`Version: ${report.meta.reasonixVersion}`);
-  console.log(`Date:    ${report.meta.date}`);
-
-  if (args.outPath) {
-    const dir = dirname(args.outPath);
-    if (dir !== ".") mkdirSync(dir, { recursive: true });
-    writeFileSync(args.outPath, `${JSON.stringify(report, null, 2)}\n`, "utf8");
-    console.log(`Wrote ${args.outPath}`);
-  }
-
-  if (passed < total) process.exit(1);
-}
-
-if (isMain()) {
-  main().catch((err) => {
-    console.error(err);
-    process.exit(1);
-  });
-}

benchmarks/behavior-stability/scenarios.ts

@@ -1,147 +0,0 @@
-/** Behavior-stability scenario for PR #1462 — constraint persistence across context folds. */
-
-import { AppendOnlyLog, SessionStats } from "../../src/index.js";
-import { ContextManager } from "../../src/context-manager.js";
-import { Usage } from "../../src/client.js";
-import type { EvalResult, EvalScenario } from "./types.js";
-
-/* ------------------------------------------------------------------ */
-/*  Helpers                                                            */
-/* ------------------------------------------------------------------ */
-
-function withTiming(fn: () => Promise<EvalResult>): Promise<EvalResult> {
-  const t0 = Date.now();
-  return fn().then((r) => ({ ...r, durationMs: Date.now() - t0 }));
-}
-
-function pass(id: string, details: string): EvalResult {
-  return { scenarioId: id, pass: true, durationMs: 0, details };
-}
-
-function fail(id: string, details: string): EvalResult {
-  return { scenarioId: id, pass: false, durationMs: 0, details };
-}
-
-/* ------------------------------------------------------------------ */
-/*  Mock client — returns a fixed summary for fold testing             */
-/* ------------------------------------------------------------------ */
-
-class MockDeepSeekClient {
-  private _replyContent: string;
-  constructor(replyContent: string) {
-    this._replyContent = replyContent;
-  }
-  async chat(): Promise<{
-    content: string;
-    reasoningContent: string;
-    usage: Usage;
-  }> {
-    return {
-      content: this._replyContent,
-      reasoningContent: "",
-      usage: new Usage(100, 50, 150, 0, 100),
-    };
-  }
-}
-
-/* ------------------------------------------------------------------ */
-/*  #1462 — Constraint persistence across folds (local)                */
-/* ------------------------------------------------------------------ */
-
-const constraintPersistence: EvalScenario = {
-  id: "constraint-persistence",
-  name: "Context fold preserves pinned constraints from system prompt",
-  category: "context",
-  requiresApi: false,
-  run: () =>
-    withTiming(async () => {
-      const systemPrompt = `You are a coding assistant.
-
-# HIGH PRIORITY constraints
-
-- DO NOT use npm install without asking.
-- Never delete files without confirmation.
-
-# User memory
-
-The user prefers TypeScript over JavaScript.
-
-# Project memory
-
-This project uses pnpm instead of npm.`;
-
-      const log = new AppendOnlyLog();
-      // Fill log with synthetic long history to trigger a fold boundary > 0.
-      for (let i = 0; i < 40; i++) {
-        log.append({
-          role: "user",
-          content: `Turn ${i}: Please review the codebase and tell me what you found.`,
-        });
-        log.append({
-          role: "assistant",
-          content: `Turn ${i} summary: I reviewed the codebase and found several issues including unused imports, missing type annotations, and inconsistent formatting.`,
-        });
-      }
-
-      const client = new MockDeepSeekClient("Summary of all prior turns.");
-      const stats = new SessionStats();
-      const ctx = new ContextManager({
-        client: client as any,
-        log,
-        stats,
-        sessionName: null,
-        getAbortSignal: () => new AbortController().signal,
-        getCurrentTurn: () => 1,
-        getSystemPrompt: () => systemPrompt,
-      });
-
-      // Use a tiny tail budget so almost everything becomes head and gets summarized.
-      const result = await ctx.fold("deepseek-v4-flash", { keepRecentTokens: 200 });
-
-      if (!result.folded) {
-        return fail(
-          "constraint-persistence",
-          `Fold did not happen (before=${result.beforeMessages}, after=${result.afterMessages}).`,
-        );
-      }
-
-      const foldedSystem = log.toMessages()[0];
-      if (foldedSystem?.role !== "assistant") {
-        return fail(
-          "constraint-persistence",
-          `Expected first folded message to be assistant summary, got ${foldedSystem?.role}.`,
-        );
-      }
-
-      const content = foldedSystem.content as string;
-      const checks = [
-        { text: "DO NOT use npm install without asking", label: "HIGH PRIORITY constraint" },
-        { text: "Never delete files without confirmation", label: "HIGH PRIORITY constraint 2" },
-        { text: "TypeScript over JavaScript", label: "User memory" },
-        { text: "pnpm instead of npm", label: "Project memory" },
-      ];
-
-      const missing = checks.filter((c) => !content.includes(c.text));
-      if (missing.length > 0) {
-        return fail(
-          "constraint-persistence",
-          `Folded summary missing ${missing.length} constraint(s): ${missing.map((m) => m.label).join(", ")}.`,
-        );
-      }
-
-      return pass(
-        "constraint-persistence",
-        `Fold reduced ${result.beforeMessages} → ${result.afterMessages} messages and preserved all ${checks.length} pinned constraints.`,
-      );
-    }),
-};
-
-/* ------------------------------------------------------------------ */
-/*  Exports                                                            */
-/* ------------------------------------------------------------------ */
-
-export const LOCAL_SCENARIOS: EvalScenario[] = [constraintPersistence];
-
-export const API_SCENARIOS: EvalScenario[] = [];
-
-export const ALL_SCENARIOS: EvalScenario[] = [...LOCAL_SCENARIOS, ...API_SCENARIOS];

benchmarks/behavior-stability/types.ts

@@ -1,26 +0,0 @@
-/** Behavior-stability evaluation types — lightweight, local-first regression suite. */
-
-export interface EvalScenario {
-  id: string;
-  name: string;
-  category: "shell" | "context" | "integrity";
-  /** True when the scenario calls the live LLM API (costs money). */
-  requiresApi: boolean;
-  run: () => Promise<EvalResult>;
-}
-
-export interface EvalResult {
-  scenarioId: string;
-  pass: boolean;
-  durationMs: number;
-  details: string;
-}
-
-export interface EvalReport {
-  meta: {
-    date: string;
-    reasonixVersion: string;
-    model?: string;
-  };
-  results: EvalResult[];
-}

benchmarks/compression-eval/.gitignore

@@ -1 +0,0 @@
-runs/

benchmarks/compression-eval/driver.ts

@@ -1,165 +0,0 @@
-import { readFileSync, writeFileSync } from "node:fs";
-import { resolvePreset } from "../../src/cli/ui/presets.js";
-import { DeepSeekClient } from "../../src/client.js";
-import { codeSystemPrompt } from "../../src/code/prompt.js";
-import { buildCodeToolset } from "../../src/code/setup.js";
-import { loadApiKey, loadBaseUrl, loadPreset } from "../../src/config.js";
-import { loadDotenv } from "../../src/env.js";
-import { CacheFirstLoop } from "../../src/loop.js";
-import { ImmutablePrefix } from "../../src/memory/runtime.js";
-import { DEEPSEEK_CONTEXT_TOKENS } from "../../src/telemetry/stats.js";
-
-interface DriverOptions {
-  taskFile: string;
-  transcriptOut: string;
-  budgetUsd: number;
-  rootDir: string;
-  /** Override DEEPSEEK_CONTEXT_TOKENS to force fold triggers on short tasks. */
-  fakeCtxMax: number | null;
-}
-
-function parseArgs(argv: string[]): DriverOptions {
-  let taskFile = "";
-  let transcriptOut = "";
-  let budgetUsd = 2.0;
-  let rootDir = process.cwd();
-  let fakeCtxMax: number | null = null;
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i]!;
-    if (a === "--task-file") taskFile = argv[++i]!;
-    else if (a === "--transcript-out") transcriptOut = argv[++i]!;
-    else if (a === "--budget") budgetUsd = Number(argv[++i]);
-    else if (a === "--root") rootDir = argv[++i]!;
-    else if (a === "--fake-ctxmax") fakeCtxMax = Number(argv[++i]);
-  }
-  if (!taskFile || !transcriptOut) {
-    console.error(
-      "usage: driver.ts --task-file <p> --transcript-out <p> [--budget 2] [--root <dir>] [--fake-ctxmax N]",
-    );
-    process.exit(1);
-  }
-  return { taskFile, transcriptOut, budgetUsd, rootDir, fakeCtxMax };
-}
-
-interface TurnRecord {
-  turn: number;
-  role: string;
-  prompt_tokens?: number;
-  cache_hit_tokens?: number;
-  cache_miss_tokens?: number;
-  cost_usd?: number;
-  tool?: string;
-  content_preview?: string;
-}
-
-async function main(): Promise<void> {
-  loadDotenv();
-  const opts = parseArgs(process.argv.slice(2));
-  const key = loadApiKey();
-  if (!key) {
-    console.error("DEEPSEEK_API_KEY missing — run `reasonix setup` first");
-    process.exit(1);
-  }
-  process.env.DEEPSEEK_API_KEY = key;
-
-  const presetSettings = resolvePreset(loadPreset());
-  const model = presetSettings.model;
-  if (opts.fakeCtxMax !== null && opts.fakeCtxMax > 0) {
-    for (const k of Object.keys(DEEPSEEK_CONTEXT_TOKENS)) {
-      DEEPSEEK_CONTEXT_TOKENS[k] = opts.fakeCtxMax;
-    }
-    console.log(`ctxMax override (ALL deepseek models): -> ${opts.fakeCtxMax}`);
-  }
-  console.log(`model: ${model}`);
-  console.log(`root:  ${opts.rootDir}`);
-  console.log(`budget: $${opts.budgetUsd}`);
-
-  const task = readFileSync(opts.taskFile, "utf8").trim();
-  console.log(`task: ${task.length} chars`);
-
-  const toolset = await buildCodeToolset({ rootDir: opts.rootDir });
-  console.log(`tools: ${toolset.tools.specs().length} registered`);
-
-  const client = new DeepSeekClient({ baseUrl: loadBaseUrl() });
-  const prefix = new ImmutablePrefix({
-    system: codeSystemPrompt(opts.rootDir),
-    toolSpecs: toolset.tools.specs(),
-  });
-  const loop = new CacheFirstLoop({
-    client,
-    prefix,
-    tools: toolset.tools,
-    model,
-    budgetUsd: opts.budgetUsd,
-  });
-
-  const records: TurnRecord[] = [];
-  const t0 = Date.now();
-  let lastTurn = 0;
-  try {
-    for await (const ev of loop.step(task)) {
-      if (ev.role === "assistant_final" && ev.stats?.usage) {
-        const u = ev.stats.usage;
-        const rec: TurnRecord = {
-          turn: ev.turn,
-          role: "assistant_final",
-          prompt_tokens: u.promptTokens,
-          cache_hit_tokens: u.promptCacheHitTokens,
-          cache_miss_tokens: u.promptCacheMissTokens,
-          cost_usd: ev.stats.cost,
-          content_preview: (ev.content ?? "").slice(0, 200),
-        };
-        records.push(rec);
-        lastTurn = ev.turn;
-        const ratio = u.promptCacheHitTokens / (u.promptTokens || 1);
-        process.stdout.write(
-          `  turn ${ev.turn}: pt=${u.promptTokens} hit=${(ratio * 100).toFixed(1)}% cost=$${ev.stats.cost.toFixed(4)}\n`,
-        );
-      } else if (ev.role === "tool" && ev.toolName) {
-        records.push({
-          turn: ev.turn,
-          role: "tool",
-          tool: ev.toolName,
-          content_preview: (ev.content ?? "").slice(0, 200),
-        });
-      } else if (ev.role === "warning" || ev.role === "status") {
-        const lc = (ev.content ?? "").toLowerCase();
-        const isFold = lc.includes("fold") || lc.includes("compact") || lc.includes("summar");
-        records.push({
-          turn: ev.turn,
-          role: isFold ? "fold_event" : `${ev.role}_event`,
-          content_preview: ev.content,
-        });
-        if (isFold) {
-          process.stdout.write(
-            `  ⟳ ${ev.role.toUpperCase()} at turn ${ev.turn}: ${ev.content.slice(0, 120)}\n`,
-          );
-        }
-      } else if (ev.role === "error") {
-        records.push({ turn: ev.turn, role: "error", content_preview: ev.error });
-        process.stdout.write(`  ✗ error turn ${ev.turn}: ${ev.error}\n`);
-      } else if (ev.role === "done") {
-        process.stdout.write(`  ✓ done (turn ${ev.turn})\n`);
-      }
-    }
-  } catch (err) {
-    process.stdout.write(`  ✗ loop threw: ${(err as Error).message}\n`);
-  }
-
-  const elapsed = Math.round((Date.now() - t0) / 1000);
-  const s = loop.stats.summary();
-  console.log(
-    `\nDONE turns=${lastTurn} elapsed=${elapsed}s cache=${(s.cacheHitRatio * 100).toFixed(1)}% cost=$${s.totalCostUsd.toFixed(4)}`,
-  );
-
-  writeFileSync(
-    opts.transcriptOut,
-    `${JSON.stringify({ _meta: { model, task_length: task.length, elapsed_s: elapsed, summary: s } })}\n${records.map((r) => JSON.stringify(r)).join("\n")}\n`,
-  );
-  console.log(`transcript: ${opts.transcriptOut}`);
-}
-
-main().catch((err) => {
-  console.error("driver crashed:", err);
-  process.exit(1);
-});

benchmarks/compression-eval/e2e.ts

@@ -1,325 +0,0 @@
-import { spawnSync } from "node:child_process";
-import { mkdirSync, readFileSync, writeFileSync } from "node:fs";
-import { join } from "node:path";
-
-interface RunSpec {
-  name: string;
-  foldThreshold: number;
-  aggressiveThreshold: number;
-}
-
-interface RunSummary {
-  name: string;
-  repeat: number;
-  turns: number;
-  cacheHitRatio: number;
-  costUsd: number;
-  totalInputTokens: number;
-  cacheHitTokens: number;
-  cacheMissTokens: number;
-  foldCount: number;
-  exitCode: number;
-}
-
-const RUNS: RunSpec[] = [
-  { name: "no-fold", foldThreshold: 0.99, aggressiveThreshold: 0.99 },
-  { name: "current_50_70", foldThreshold: 0.5, aggressiveThreshold: 0.7 },
-  { name: "very_late_75_90", foldThreshold: 0.75, aggressiveThreshold: 0.9 },
-  { name: "aggressive_90_95", foldThreshold: 0.9, aggressiveThreshold: 0.95 },
-];
-
-function parseArgs(argv: string[]): {
-  taskFile: string;
-  budget: number;
-  outDir: string;
-  only: string | null;
-  fakeCtxMax: number | null;
-  repeats: number;
-} {
-  let taskFile = "";
-  let budget = 2.0;
-  let outDir = "";
-  let only: string | null = null;
-  let fakeCtxMax: number | null = null;
-  let repeats = 1;
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i]!;
-    if (a === "--task-file") taskFile = argv[++i]!;
-    else if (a === "--budget") budget = Number(argv[++i]);
-    else if (a === "--out-dir") outDir = argv[++i]!;
-    else if (a === "--only") only = argv[++i]!;
-    else if (a === "--fake-ctxmax") fakeCtxMax = Number(argv[++i]);
-    else if (a === "--repeats") repeats = Number(argv[++i]);
-  }
-  if (!taskFile) {
-    console.error(
-      "usage: tsx e2e.ts --task-file <p> [--budget N] [--out-dir <p>] [--only <n>] [--fake-ctxmax N] [--repeats N]",
-    );
-    process.exit(1);
-  }
-  if (!outDir) {
-    const ts = new Date().toISOString().replace(/[:.]/g, "-").slice(0, 19);
-    outDir = `benchmarks/compression-eval/runs/${ts}`;
-  }
-  return { taskFile, budget, outDir, only, fakeCtxMax, repeats };
-}
-
-function loadTranscriptSummary(path: string): Partial<RunSummary> {
-  let turns = 0;
-  let inputTokens = 0;
-  let cacheHit = 0;
-  let cacheMiss = 0;
-  let costUsd = 0;
-  let foldCount = 0;
-  const promptSeq: number[] = [];
-  try {
-    const lines = readFileSync(path, "utf8").split("\n");
-    for (const ln of lines) {
-      const t = ln.trim();
-      if (!t) continue;
-      let rec: {
-        role?: string;
-        prompt_tokens?: number;
-        cache_hit_tokens?: number;
-        cache_miss_tokens?: number;
-        cost_usd?: number;
-      };
-      try {
-        rec = JSON.parse(t);
-      } catch {
-        continue;
-      }
-      if (rec.role === "assistant_final") {
-        turns++;
-        const pt = rec.prompt_tokens ?? 0;
-        inputTokens += pt;
-        cacheHit += rec.cache_hit_tokens ?? 0;
-        cacheMiss += rec.cache_miss_tokens ?? 0;
-        costUsd += rec.cost_usd ?? 0;
-        if (pt > 0) promptSeq.push(pt);
-      } else if (rec.role === "fold_event") {
-        foldCount++;
-      }
-    }
-  } catch {
-    /* file missing — caller will see exitCode and zeros */
-  }
-  if (foldCount === 0) {
-    for (let i = 1; i < promptSeq.length; i++) {
-      if (promptSeq[i]! < promptSeq[i - 1]! * 0.6) foldCount++;
-    }
-  }
-  return {
-    turns,
-    totalInputTokens: inputTokens,
-    cacheHitTokens: cacheHit,
-    cacheMissTokens: cacheMiss,
-    cacheHitRatio: inputTokens > 0 ? cacheHit / inputTokens : 0,
-    costUsd,
-    foldCount,
-  };
-}
-
-function runOne(
-  spec: RunSpec,
-  task: string,
-  budget: number,
-  outDir: string,
-  fakeCtxMax: number | null,
-  repeat: number,
-): RunSummary {
-  const runName = repeat > 1 ? `${spec.name}_r${repeat}` : spec.name;
-  const runDir = join(outDir, runName);
-  mkdirSync(runDir, { recursive: true });
-  const transcript = join(runDir, "transcript.jsonl");
-  const logPath = join(runDir, "stdout.log");
-
-  console.log(
-    `\n>>> [${spec.name}] fold=${spec.foldThreshold} agg=${spec.aggressiveThreshold}`,
-  );
-  const tsxCli = join("node_modules", "tsx", "dist", "cli.mjs");
-  const taskFilePath = join(runDir, "task.txt");
-  writeFileSync(taskFilePath, task);
-  const t0 = Date.now();
-  const res = spawnSync(
-    process.execPath,
-    [
-      tsxCli,
-      "benchmarks/compression-eval/driver.ts",
-      "--task-file",
-      taskFilePath,
-      "--transcript-out",
-      transcript,
-      "--budget",
-      String(budget),
-      ...(fakeCtxMax ? ["--fake-ctxmax", String(fakeCtxMax)] : []),
-    ],
-    {
-      env: {
-        ...process.env,
-        REASONIX_FOLD_THRESHOLD: String(spec.foldThreshold),
-        REASONIX_FOLD_AGGRESSIVE_THRESHOLD: String(spec.aggressiveThreshold),
-      },
-      encoding: "utf8",
-      stdio: ["ignore", "pipe", "pipe"],
-      shell: false,
-      maxBuffer: 1024 * 1024 * 128,
-    },
-  );
-  const exitCode = res.status ?? -1;
-  const stdout = res.stdout ?? "";
-  const stderr = res.stderr ?? "";
-  const elapsedS = Math.round((Date.now() - t0) / 1000);
-  writeFileSync(logPath, `${stdout}\n--- stderr ---\n${stderr}`);
-  console.log(`    exit=${exitCode} elapsed=${elapsedS}s log=${logPath}`);
-
-  const parsed = loadTranscriptSummary(transcript);
-  return {
-    name: spec.name,
-    repeat,
-    turns: parsed.turns ?? 0,
-    cacheHitRatio: parsed.cacheHitRatio ?? 0,
-    costUsd: parsed.costUsd ?? 0,
-    totalInputTokens: parsed.totalInputTokens ?? 0,
-    cacheHitTokens: parsed.cacheHitTokens ?? 0,
-    cacheMissTokens: parsed.cacheMissTokens ?? 0,
-    foldCount: parsed.foldCount ?? 0,
-    exitCode,
-  };
-}
-
-function pct(n: number): string {
-  return `${(n * 100).toFixed(1)}%`;
-}
-function fmtTok(n: number): string {
-  if (n >= 1e6) return `${(n / 1e6).toFixed(2)}M`;
-  if (n >= 1e3) return `${(n / 1e3).toFixed(1)}k`;
-  return String(n);
-}
-
-interface AggregateRow {
-  name: string;
-  n: number;
-  meanCost: number;
-  stdCost: number;
-  meanMiss: number;
-  stdMiss: number;
-  meanHitPct: number;
-  stdHitPct: number;
-  meanTurns: number;
-  meanFolds: number;
-}
-
-function meanStd(xs: number[]): [number, number] {
-  if (xs.length === 0) return [0, 0];
-  const m = xs.reduce((a, b) => a + b, 0) / xs.length;
-  const v = xs.reduce((a, b) => a + (b - m) ** 2, 0) / xs.length;
-  return [m, Math.sqrt(v)];
-}
-
-function aggregateByConfig(rows: RunSummary[]): AggregateRow[] {
-  const groups = new Map<string, RunSummary[]>();
-  for (const r of rows) {
-    if (!groups.has(r.name)) groups.set(r.name, []);
-    groups.get(r.name)!.push(r);
-  }
-  const result: AggregateRow[] = [];
-  for (const [name, group] of groups) {
-    const [meanCost, stdCost] = meanStd(group.map((r) => r.costUsd));
-    const [meanMiss, stdMiss] = meanStd(group.map((r) => r.cacheMissTokens));
-    const [meanHitPct, stdHitPct] = meanStd(group.map((r) => r.cacheHitRatio * 100));
-    const [meanTurns] = meanStd(group.map((r) => r.turns));
-    const [meanFolds] = meanStd(group.map((r) => r.foldCount));
-    result.push({
-      name,
-      n: group.length,
-      meanCost,
-      stdCost,
-      meanMiss,
-      stdMiss,
-      meanHitPct,
-      stdHitPct,
-      meanTurns,
-      meanFolds,
-    });
-  }
-  return result;
-}
-
-function reportAggregate(rows: AggregateRow[]): string {
-  const cols: Array<[string, (r: AggregateRow) => string]> = [
-    ["config", (r) => r.name],
-    ["n", (r) => String(r.n)],
-    ["turns", (r) => r.meanTurns.toFixed(1)],
-    ["folds", (r) => r.meanFolds.toFixed(1)],
-    ["hit%", (r) => `${r.meanHitPct.toFixed(1)}±${r.stdHitPct.toFixed(1)}`],
-    ["miss", (r) => `${fmtTok(r.meanMiss)}±${fmtTok(r.stdMiss)}`],
-    ["cost$", (r) => `${r.meanCost.toFixed(4)}±${r.stdCost.toFixed(4)}`],
-  ];
-  const widths = cols.map(([h, f]) => Math.max(h.length, ...rows.map((r) => f(r).length)));
-  const head = cols.map(([h], i) => h.padEnd(widths[i]!)).join("│ ");
-  const sep = "─".repeat(widths.reduce((a, b) => a + b + 2, 0));
-  const body = rows
-    .map((r) => cols.map(([_, f], i) => f(r).padEnd(widths[i]!)).join("│ "))
-    .join("\n");
-  return `${head}\n${sep}\n${body}`;
-}
-
-function reportTable(rows: RunSummary[]): string {
-  const cols: Array<[string, (r: RunSummary) => string]> = [
-    ["config", (r) => r.name],
-    ["turns", (r) => String(r.turns)],
-    ["input", (r) => fmtTok(r.totalInputTokens)],
-    ["hit", (r) => fmtTok(r.cacheHitTokens)],
-    ["hit%", (r) => pct(r.cacheHitRatio)],
-    ["miss", (r) => fmtTok(r.cacheMissTokens)],
-    ["folds", (r) => String(r.foldCount)],
-    ["cost", (r) => `$${r.costUsd.toFixed(4)}`],
-    ["exit", (r) => String(r.exitCode)],
-  ];
-  const widths = cols.map(([h, f]) => Math.max(h.length, ...rows.map((r) => f(r).length)));
-  const head = cols.map(([h], i) => h.padEnd(widths[i]!)).join("│ ");
-  const sep = "─".repeat(widths.reduce((a, b) => a + b + 2, 0));
-  const body = rows
-    .map((r) => cols.map(([_, f], i) => f(r).padEnd(widths[i]!)).join("│ "))
-    .join("\n");
-  return `${head}\n${sep}\n${body}`;
-}
-
-function main(): void {
-  const { taskFile, budget, outDir, only, fakeCtxMax, repeats } = parseArgs(process.argv.slice(2));
-  const task = readFileSync(taskFile, "utf8").trim();
-  mkdirSync(outDir, { recursive: true });
-  const specs = only ? RUNS.filter((r) => r.name === only) : RUNS;
-  if (specs.length === 0) {
-    console.error(`no config matches --only=${only}; available: ${RUNS.map((r) => r.name).join(", ")}`);
-    process.exit(1);
-  }
-  console.log(`task: ${taskFile} (${task.length} chars)`);
-  console.log(`out:  ${outDir}`);
-  console.log(`budget per run: $${budget}`);
-  console.log(`configs: ${specs.map((s) => s.name).join(", ")}`);
-  console.log(`repeats: ${repeats}`);
-  if (fakeCtxMax) console.log(`fake ctxMax: ${fakeCtxMax}`);
-
-  const summaries: RunSummary[] = [];
-  for (let r = 1; r <= repeats; r++) {
-    for (const spec of specs) {
-      summaries.push(runOne(spec, task, budget, outDir, fakeCtxMax, r));
-    }
-  }
-
-  const table = reportTable(summaries);
-  console.log(`\n=== Raw runs ===\n${table}`);
-  if (repeats > 1) {
-    const agg = aggregateByConfig(summaries);
-    const aggTable = reportAggregate(agg);
-    console.log(`\n=== Aggregate (mean ± std) ===\n${aggTable}`);
-    writeFileSync(join(outDir, "aggregate.txt"), aggTable);
-  }
-  writeFileSync(join(outDir, "summary.json"), JSON.stringify(summaries, null, 2));
-  writeFileSync(join(outDir, "summary.txt"), table);
-  console.log(`\nsaved: ${join(outDir, "summary.json")}`);
-}
-
-main();

benchmarks/compression-eval/run.ts

@@ -1,89 +0,0 @@
-import { basename } from "node:path";
-import { loadSession, simulate, type RunResult, type ThresholdConfig } from "./simulator.js";
-
-const BASELINE: ThresholdConfig = {
-  ctxMax: 1_000_000,
-  foldThreshold: 0.5,
-  tailFraction: 0.2,
-  aggressiveThreshold: 0.7,
-  aggressiveTailFraction: 0.1,
-  minSavingsFraction: 0.3,
-  summaryRatio: 0.05,
-};
-
-interface NamedConfig {
-  name: string;
-  config: ThresholdConfig;
-}
-
-function configs(ctxMax: number): NamedConfig[] {
-  const mk = (over: Partial<ThresholdConfig>): ThresholdConfig => ({ ...BASELINE, ...over, ctxMax });
-  return [
-    { name: "no-fold", config: mk({ foldThreshold: 999, aggressiveThreshold: 999 }) },
-    { name: "current (50/70)", config: mk({ foldThreshold: 0.5, aggressiveThreshold: 0.7 }) },
-    { name: "late (60/80)", config: mk({ foldThreshold: 0.6, aggressiveThreshold: 0.8 }) },
-    { name: "very-late (75/90)", config: mk({ foldThreshold: 0.75, aggressiveThreshold: 0.9 }) },
-    { name: "early (35/55)", config: mk({ foldThreshold: 0.35, aggressiveThreshold: 0.55 }) },
-  ];
-}
-
-function parseArgs(argv: string[]): { sessions: string[]; ctxMax: number } {
-  const sessions: string[] = [];
-  let ctxMax = 1_000_000;
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i]!;
-    if (a === "--ctx-max") ctxMax = Number(argv[++i]);
-    else if (a === "--session") sessions.push(argv[++i]!);
-    else if (!a.startsWith("--")) sessions.push(a);
-  }
-  return { sessions, ctxMax };
-}
-
-function fmt(n: number): string {
-  if (n >= 1e9) return `${(n / 1e9).toFixed(2)}G`;
-  if (n >= 1e6) return `${(n / 1e6).toFixed(2)}M`;
-  if (n >= 1e3) return `${(n / 1e3).toFixed(1)}k`;
-  return String(n);
-}
-
-function pct(n: number): string {
-  return `${(n * 100).toFixed(1)}%`;
-}
-
-function report(label: string, rows: Array<{ name: string; result: RunResult }>): void {
-  console.log(`\n=== ${label} ===`);
-  const w = [22, 10, 10, 8, 10, 6, 10];
-  const headers = ["config", "input", "cache_hit", "hit%", "miss", "folds", "peak"];
-  console.log(headers.map((h, i) => h.padEnd(w[i]!)).join("│ "));
-  console.log("─".repeat(w.reduce((a, b) => a + b + 2, 0)));
-  for (const { name, result: r } of rows) {
-    const peak = r.perTurn.reduce((m, t) => Math.max(m, t.ratio), 0);
-    console.log(
-      [
-        name.padEnd(w[0]!),
-        fmt(r.totalInputTokens).padEnd(w[1]!),
-        fmt(r.totalCacheHitTokens).padEnd(w[2]!),
-        pct(r.cacheHitRatio).padEnd(w[3]!),
-        fmt(r.totalCacheMissTokens).padEnd(w[4]!),
-        String(r.foldCount).padEnd(w[5]!),
-        pct(peak).padEnd(w[6]!),
-      ].join("│ "),
-    );
-  }
-}
-
-function main(): void {
-  const { sessions, ctxMax } = parseArgs(process.argv.slice(2));
-  if (sessions.length === 0) {
-    console.error("usage: tsx run.ts [--ctx-max N] <session.jsonl> [more...]");
-    process.exit(1);
-  }
-  for (const path of sessions) {
-    const msgs = loadSession(path);
-    const cfgs = configs(ctxMax);
-    const rows = cfgs.map(({ name, config }) => ({ name, result: simulate(msgs, config, path) }));
-    report(`${basename(path)}  (${msgs.length} msgs, ctxMax=${fmt(ctxMax)})`, rows);
-  }
-}
-
-main();

benchmarks/compression-eval/simulator.ts

@@ -1,197 +0,0 @@
-import { readFileSync } from "node:fs";
-import {
-  countTokensBounded,
-  estimateConversationTokens,
-  estimateRequestTokens,
-} from "../../src/tokenizer.js";
-import type { ChatMessage } from "../../src/types.js";
-
-export interface ThresholdConfig {
-  ctxMax: number;
-  foldThreshold: number;
-  tailFraction: number;
-  aggressiveThreshold: number;
-  aggressiveTailFraction: number;
-  minSavingsFraction: number;
-  summaryRatio: number;
-}
-
-export interface FoldEvent {
-  turn: number;
-  ratioBeforeFold: number;
-  promptTokensBefore: number;
-  headMessages: number;
-  tailMessages: number;
-  headTokens: number;
-  summaryTokens: number;
-  aggressive: boolean;
-}
-
-export interface TurnRecord {
-  turn: number;
-  promptTokens: number;
-  cacheHitTokens: number;
-  cacheMissTokens: number;
-  ratio: number;
-  folded: boolean;
-}
-
-export interface RunResult {
-  config: ThresholdConfig;
-  sessionPath: string;
-  totalTurns: number;
-  totalInputTokens: number;
-  totalCacheHitTokens: number;
-  totalCacheMissTokens: number;
-  cacheHitRatio: number;
-  foldCount: number;
-  folds: FoldEvent[];
-  perTurn: TurnRecord[];
-}
-
-export function loadSession(path: string): ChatMessage[] {
-  const lines = readFileSync(path, "utf8").split("\n");
-  const out: ChatMessage[] = [];
-  for (const ln of lines) {
-    const t = ln.trim();
-    if (!t) continue;
-    try {
-      out.push(JSON.parse(t) as ChatMessage);
-    } catch {
-      continue;
-    }
-  }
-  return out;
-}
-
-function messageTokens(m: ChatMessage): number {
-  return estimateConversationTokens([m], true);
-}
-
-function messageKey(m: ChatMessage): string {
-  return JSON.stringify({
-    r: m.role,
-    c: m.content ?? "",
-    tc: m.tool_calls ?? null,
-    tcid: m.tool_call_id ?? null,
-  });
-}
-
-function commonPrefixTokens(curr: ChatMessage[], prev: ChatMessage[]): number {
-  const n = Math.min(curr.length, prev.length);
-  let tokens = 0;
-  for (let i = 0; i < n; i++) {
-    if (messageKey(curr[i]!) !== messageKey(prev[i]!)) break;
-    tokens += messageTokens(curr[i]!);
-  }
-  return tokens;
-}
-
-function findFoldBoundary(log: ChatMessage[], tailBudget: number): number {
-  let cum = 0;
-  let boundary = log.length;
-  for (let i = log.length - 1; i >= 0; i--) {
-    const t = messageTokens(log[i]!);
-    if (cum + t > tailBudget) break;
-    cum += t;
-    if (log[i]!.role === "user") boundary = i;
-  }
-  return boundary;
-}
-
-function synthesizeSummary(head: ChatMessage[], targetTokens: number): ChatMessage {
-  const userBits: string[] = [];
-  for (const m of head) {
-    if (m.role === "user" && typeof m.content === "string") userBits.push(m.content);
-  }
-  const seed = userBits.join("\n") || "earlier turns folded";
-  let text = `[CONVERSATION HISTORY SUMMARY — fold of ${head.length} messages]\n\n${seed}`;
-  const currentTokens = countTokensBounded(text);
-  if (currentTokens > targetTokens) {
-    const ratio = targetTokens / currentTokens;
-    text = text.slice(0, Math.max(80, Math.floor(text.length * ratio)));
-  }
-  return { role: "assistant", content: text, reasoning_content: "" };
-}
-
-function turnBoundaries(messages: ChatMessage[]): number[] {
-  const ends: number[] = [];
-  for (let i = 0; i < messages.length; i++) {
-    if (messages[i]!.role === "assistant") ends.push(i);
-  }
-  return ends;
-}
-
-export function simulate(messages: ChatMessage[], config: ThresholdConfig, sessionPath: string): RunResult {
-  const ends = turnBoundaries(messages);
-  const log: ChatMessage[] = [];
-  let prevSnapshot: ChatMessage[] = [];
-  const perTurn: TurnRecord[] = [];
-  const folds: FoldEvent[] = [];
-  let totalInput = 0;
-  let totalHit = 0;
-  let totalMiss = 0;
-  let cursor = 0;
-
-  for (let turn = 0; turn < ends.length; turn++) {
-    const end = ends[turn]!;
-    for (let i = cursor; i < end; i++) log.push(messages[i]!);
-    cursor = end;
-
-    const promptTokens = estimateRequestTokens(log, null, true);
-    const ratio = promptTokens / config.ctxMax;
-    const hit = commonPrefixTokens(log, prevSnapshot);
-    const miss = Math.max(0, promptTokens - hit);
-    let folded = false;
-
-    if (ratio > config.foldThreshold) {
-      const aggressive = ratio > config.aggressiveThreshold;
-      const tailFrac = aggressive ? config.aggressiveTailFraction : config.tailFraction;
-      const tailBudget = Math.floor(config.ctxMax * tailFrac);
-      const boundary = findFoldBoundary(log, tailBudget);
-      if (boundary > 0) {
-        const head = log.slice(0, boundary);
-        const tail = log.slice(boundary);
-        const headTokens = head.reduce((a, m) => a + messageTokens(m), 0);
-        const totalTokens = log.reduce((a, m) => a + messageTokens(m), 0);
-        if (headTokens >= totalTokens * config.minSavingsFraction) {
-          const summaryTokens = Math.max(64, Math.floor(headTokens * config.summaryRatio));
-          const summary = synthesizeSummary(head, summaryTokens);
-          const next = [summary, ...tail];
-          log.length = 0;
-          for (const m of next) log.push(m);
-          folds.push({
-            turn,
-            ratioBeforeFold: ratio,
-            promptTokensBefore: promptTokens,
-            headMessages: head.length,
-            tailMessages: tail.length,
-            headTokens,
-            summaryTokens,
-            aggressive,
-          });
-          folded = true;
-        }
-      }
-    }
-
-    totalInput += promptTokens;
-    totalHit += hit;
-    totalMiss += miss;
-    perTurn.push({ turn, promptTokens, cacheHitTokens: hit, cacheMissTokens: miss, ratio, folded });
-    prevSnapshot = log.slice();
-  }
-
-  return {
-    config,
-    sessionPath,
-    totalTurns: ends.length,
-    totalInputTokens: totalInput,
-    totalCacheHitTokens: totalHit,
-    totalCacheMissTokens: totalMiss,
-    cacheHitRatio: totalInput > 0 ? totalHit / totalInput : 0,
-    foldCount: folds.length,
-    folds,
-    perTurn,
-  };
-}

benchmarks/compression-eval/task.txt

@@ -1,13 +0,0 @@
-为这个仓库的 src/ 目录写一份完整的架构总览文档。
-
-要求：
-1. 用 read_file / directory_tree / grep_search 等工具，**逐个**读取 src/ 下每个子目录（cli/, mcp/, memory/, telemetry/, transcript/, tools/, ui/, i18n/, loop/ 等）的代表性文件。每个子目录至少读 3-5 个核心 .ts 文件。
-2. 不要只 grep 关键词就下结论 —— 必须打开文件读真实内容，从代码里抽出"这个模块的职责 / 它的核心数据结构 / 它对外暴露什么"。
-3. 边读边在心里搭建依赖图。
-4. 全部读完后，输出一份 markdown 报告，包含：
-   - 整体架构图（mermaid 或 ASCII，模块之间的依赖关系）
-   - 按模块列出：职责 / 关键文件 / 对外接口 / 内部依赖
-   - 至少 8 个具体的代码引用（路径:行号），证明你真读了
-   - 结尾给出"如果新人加入项目，应该按什么顺序读代码"的建议
-
-这是一个真实的代码审计任务，请不要偷懒、不要瞎编、不要用通用模板。文件读得不够多就直接说"我读得不够"。

benchmarks/e2e/tasks/compaction/task.toml

@@ -0,0 +1,3 @@
+prompt = "Begin by reading the file story/chapter-1.md. Each chapter ends by telling you which chapter to read next; follow that trail one chapter at a time — read a chapter, find where it points you, then read that next chapter — until a chapter says the trail ends. You cannot know the next chapter without reading the current one, so do not guess or read ahead. In chapter 1 a river village is named; in the final chapter on the trail a noble House is named. When the trail ends, write a file answer.txt whose entire contents are the village name from chapter 1, a single hyphen, and the House name from the final chapter — for example Riverton-Stark — and nothing else."
+max_steps = 40
+timeout_sec = 900

benchmarks/e2e/tasks/compaction/verify.sh

@@ -0,0 +1,4 @@
+set -e
+got=$(tr -d '[:space:]' < answer.txt | tr '[:upper:]' '[:lower:]')
+want="aldermoor-verrin"
+[ "$got" = "$want" ] || { echo "answer.txt normalized to '$got', want '$want'"; exit 1; }

benchmarks/e2e/tasks/compaction/workdir/story/chapter-1.md

@@ -0,0 +1,220 @@
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 4 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 10 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 16 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 22 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 28 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 34 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 40 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 46 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 52 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 58 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 64 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 70 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 76 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 82 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 88 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 94 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 100 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 106 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+It was from Aldermoor, the cluster of cottages where the north and south rivers braid together, that our courier first set out.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 112 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 118 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 124 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 130 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 136 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 142 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 148 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 154 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 160 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 166 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 172 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 178 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 184 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 190 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 196 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 202 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 208 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+To pick up the trail again, the next page you must read is chapter-4.md.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 214 of chapter 1: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 220 of chapter 1: the chronicle continues, patient and unhurried, toward its end.

benchmarks/e2e/tasks/compaction/workdir/story/chapter-2.md

@@ -0,0 +1,220 @@
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 3 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 9 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 15 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 21 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 27 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 33 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 39 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 45 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 51 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 57 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 63 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 69 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 75 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 81 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 87 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 93 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 99 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 105 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 111 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 117 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 123 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 129 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 135 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 141 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 147 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 153 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 159 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 165 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 171 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 177 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 183 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 189 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 195 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 201 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 207 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+To pick up the trail again, the next page you must read is chapter-5.md.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 213 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 219 of chapter 2: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.

benchmarks/e2e/tasks/compaction/workdir/story/chapter-3.md

@@ -0,0 +1,220 @@
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 2 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 8 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 14 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 20 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 26 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 32 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 38 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 44 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 50 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 56 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 62 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 68 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 74 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 80 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 86 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 92 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 98 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 104 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 110 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 116 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 122 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 128 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 134 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 140 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 146 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 152 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 158 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 164 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 170 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 176 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 182 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 188 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 194 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 200 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 206 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+To pick up the trail again, the next page you must read is chapter-6.md.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 212 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 218 of chapter 3: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.

benchmarks/e2e/tasks/compaction/workdir/story/chapter-4.md

@@ -0,0 +1,220 @@
+Line 1 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 7 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 13 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 19 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 25 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 31 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 37 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 43 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 49 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 55 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 61 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 67 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 73 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 79 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 85 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 91 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 97 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 103 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 109 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 115 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 121 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 127 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 133 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 139 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 145 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 151 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 157 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 163 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 169 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 175 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 181 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 187 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 193 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 199 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 205 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+To pick up the trail again, the next page you must read is chapter-2.md.
+Line 211 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 217 of chapter 4: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.

benchmarks/e2e/tasks/compaction/workdir/story/chapter-5.md

@@ -0,0 +1,220 @@
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 6 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 12 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 18 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 24 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 30 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 36 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 42 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 48 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 54 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 60 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 66 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 72 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 78 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 84 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 90 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 96 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 102 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 108 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 114 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 120 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 126 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 132 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 138 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 144 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 150 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 156 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 162 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 168 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 174 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 180 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 186 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 192 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 198 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 204 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+To pick up the trail again, the next page you must read is chapter-3.md.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 216 of chapter 5: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.

benchmarks/e2e/tasks/compaction/workdir/story/chapter-6.md

@@ -0,0 +1,220 @@
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 5 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 11 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 17 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 23 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 29 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 35 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 41 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 47 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 53 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 59 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 65 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 71 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 77 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 83 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 89 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 95 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 101 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 107 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+With that, the ancestral seal of the Verrin was pressed into the warm wax and then, before the whole assembly, ceremonially broken.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 113 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 119 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 125 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 131 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 137 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 143 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 149 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 155 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 161 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 167 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 173 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 179 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 185 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 191 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 197 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 203 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 209 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+Here the trail ends; there is no further page to follow. You now hold both names you were sent to find: the river village of Aldermoor, where the courier began in the first chapter, and the House whose seal was broken above. Your task is complete the moment you write a file named answer.txt whose entire contents are that village name, a single hyphen, and that House name, and nothing else. Write it now.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.
+Line 215 of chapter 6: the chronicle continues, patient and unhurried, toward its end.
+The road wound on through the grey hills, and the company spoke little of what lay behind them.
+Rain had come in the night, so the morning was washed and bright, and the carts ran easily.
+Old stories were traded at the fire, half-remembered and half-invented, as such stories are.
+A merchant counted his coin twice and still mistrusted the sum, for the season had been lean.
+Banners of the lesser lords hung limp in the still air above the gatehouse, faded by sun.

benchmarks/e2e/tasks/fix-add-bug/task.toml

@@ -0,0 +1,3 @@
+prompt = "The file calc.py has a bug: add(a, b) returns the wrong result. Fix add so it returns the sum of a and b. Keep the other functions unchanged."
+max_steps = 12
+timeout_sec = 180

benchmarks/e2e/tasks/fix-add-bug/verify.sh

@@ -0,0 +1,7 @@
+set -e
+python3 - <<'PY'
+import calc
+assert calc.add(2, 3) == 5, calc.add(2, 3)
+assert calc.add(10, 5) == 15, calc.add(10, 5)
+assert calc.mul(2, 3) == 6, calc.mul(2, 3)
+PY

benchmarks/e2e/tasks/fix-add-bug/workdir/calc.py

@@ -0,0 +1,6 @@
+def add(a, b):
+    return a - b
+
+
+def mul(a, b):
+    return a * b

benchmarks/e2e/tasks/fizzbuzz/task.toml

@@ -0,0 +1,3 @@
+prompt = "Create a file named fizzbuzz.py containing a function fizzbuzz(n) that returns the string 'Fizz' when n is divisible by 3, 'Buzz' when divisible by 5, 'FizzBuzz' when divisible by both 3 and 5, and otherwise the number as a string. Do not print anything at import time."
+max_steps = 12
+timeout_sec = 180

benchmarks/e2e/tasks/fizzbuzz/verify.sh

@@ -0,0 +1,8 @@
+set -e
+python3 - <<'PY'
+import fizzbuzz
+assert fizzbuzz.fizzbuzz(3) == "Fizz", fizzbuzz.fizzbuzz(3)
+assert fizzbuzz.fizzbuzz(5) == "Buzz", fizzbuzz.fizzbuzz(5)
+assert fizzbuzz.fizzbuzz(15) == "FizzBuzz", fizzbuzz.fizzbuzz(15)
+assert fizzbuzz.fizzbuzz(7) == "7", fizzbuzz.fizzbuzz(7)
+PY

benchmarks/e2e/tasks/palindrome/task.toml

@@ -0,0 +1,3 @@
+prompt = "Create a file named palindrome.py with a function is_palindrome(s) that returns True if s reads the same forwards and backwards ignoring case, spaces, and punctuation, and False otherwise. Do not print anything at import time."
+max_steps = 12
+timeout_sec = 180

benchmarks/e2e/tasks/palindrome/verify.sh

@@ -0,0 +1,7 @@
+set -e
+python3 - <<'PY'
+import palindrome
+assert palindrome.is_palindrome("Race car") is True
+assert palindrome.is_palindrome("A man, a plan, a canal: Panama") is True
+assert palindrome.is_palindrome("hello") is False
+PY

benchmarks/real-world-cache/README.md

@@ -1,77 +0,0 @@
-# Real-world cache hit — single user, single day
-
-A real Reasonix user shared their DeepSeek dashboard for **2026-05-01**.
-Used with permission, anonymized.
-
-![DeepSeek usage dashboard, 2026-05-01](2026-05-01-deepseek-dashboard.png)
-
-## The numbers
-
-| | Tokens |
-|---|---:|
-| Input — cache hit | 435,033,856 |
-| Input — cache miss | 767,616 |
-| Output | 179,763 |
-| **Day total** | **435,981,235** |
-
-**Cache hit ratio (input):**
-`435,033,856 / (435,033,856 + 767,616)` = **99.82%**
-
-## Cost — using the prices Reasonix bills against (`src/telemetry/stats.ts`)
-
-USD per 1M tokens — `inputCacheHit / inputCacheMiss / output`:
-- `deepseek-v4-flash` — `0.0028 / 0.14 / 0.28`
-- `deepseek-v4-pro` — `0.003625 / 0.435 / 0.87`
-
-Assuming **v4-flash** (the project default):
-
-| | This user (99.82% hit) | Same workload, **0% cache** |
-|---|---:|---:|
-| Cache-hit input | $1.22 | — |
-| Cache-miss input | $0.11 | $61.01 |
-| Output | $0.05 | $0.05 |
-| **Total / day** | **$1.38** | **$61.06** |
-
-→ Cache saved this user **$59.68**, or **~97.7%** off the un-cached baseline, on a single day.
-
-On **v4-pro** (6.25% of uncached input cost) the same workload would cost
-**~$2.07** vs **~$189.73** without cache — a **~98.9%** saving.
-
-## "Isn't that just DeepSeek's prefix cache?"
-
-DeepSeek's API ships prefix caching enabled by default; the *cache* is theirs,
-the *hit rate* is the client's. Same API, different clients, very different
-hit rates:
-
-- DeepSeek's own web chat: 60–80% within a single conversation, drops to 0%
-  on a new session (system prompt may differ).
-- Cherry Studio / Open WebUI / generic OpenAI-shape SDKs: typically 30–60%
-  on long sessions — history gets reordered, tool specs get re-serialized,
-  every drift breaks the prefix.
-- Cline / Continue and other XML-tool-call clients: lower still — every tool
-  result inlines into the conversation, shifting bytes the cache keys on.
-
-99.82% is what falls out of these four design choices in Reasonix:
-
-1. **`ImmutablePrefix`** (`src/memory.ts`) — system prompt + tool specs are
-   frozen at session start. Same byte sequence every turn.
-2. **`AppendOnlyLog`** — turns only append. No reorder, no edit-in-place.
-3. **`VolatileScratch`** — chain-of-thought / per-turn scratch lives outside
-   the cached prefix so it never poisons the next hit.
-4. **Auto-compact** — when context approaches the cap, older turns fold into
-   a summary message. The summary request is shaped to reuse the main agent's
-   already-cached system/tool/history prefix; the following main-agent request
-   still pays a cold miss for the newly synthesized summary segment.
-
-DeepSeek gave us cacheable bytes. The four mechanisms above are how we keep
-the bytes cacheable.
-
-## Reproduce
-
-The synthetic side of this lives in `benchmarks/tau-bench/` — same task set
-run through `CacheFirstLoop` vs a deliberately cache-hostile baseline. The
-real-world data above is what the synthetic numbers look like once a user
-runs the harness in anger.
-
-Submit your own dashboard screenshot if you want it anonymized and added
-here — open an issue.

benchmarks/spike-mcp-reconnect/results.md

@@ -1,63 +0,0 @@
-# MCP reconnect — empirical cache-prefix spike
-
-Live `deepseek-chat` (DeepSeek prefix cache enabled by default).
-System prompt: 1546 chars (~390 tokens). 5 turns each with a small
-user message; tool-set varies between turns to simulate the drift
-shapes a `/mcp reconnect <name>` would emit.
-
-## Run
-
-```
-turn                                      prompt     hit    miss    hit%      ms
---------------------------------------------------------------------------------
-1 · cold start (toolset A)                   758     640     118   84.4%    1092
-2 · same prefix (toolset A)                  753     640     113   85.0%    1535
-3 · drift: ADDED tool (toolset A+)           810     768      42   94.8%    1048
-4 · same prefix again (toolset A+)           807     768      39   95.2%    1480
-5 · drift: EDITED desc (toolset A')          761     640     121   84.1%     791
-```
-
-(Turn 1's "cold" is misleading — the prefix had been seen by the
-remote cache from an earlier run within the cache TTL.)
-
-## Findings
-
-DeepSeek's prefix cache works at chunk granularity (consistent with
-publicly documented ~128-token chunks). Three concrete lessons:
-
-1. **Append-only drift is nearly free.** Turn 3 adds one tool *at the
-   end* of the tool list — every cache chunk before the new tool
-   stays valid, only the appended bytes miss. Net: 94.8% hit, even
-   higher than the no-drift baseline (because the system prompt +
-   whole toolset-A is still cached, and the appended chunk is now
-   cached too).
-2. **Mid-stream drift loses everything past the divergence.** Turn 5
-   edits a description on the *first* tool, so divergence falls
-   inside the tools block early. Hit drops to 84.1% — still high
-   here only because the system prompt occupies enough chunks before
-   the divergence point.
-3. **Position of the drift dominates the cost.** A trailing addition
-   is essentially zero. An edit near the start of tools is more
-   expensive. An edit in the system prompt itself (not tested) would
-   wipe the cache to zero — expected, but irrelevant for reconnect
-   since we don't change the system prompt on reconnect.
-
-## Implication for RFC #110
-
-The "any drift = full cache miss" framing in the RFC body is too
-pessimistic. The real cost of accepting a drifted reconnect depends
-on *where* the drift lands:
-
-- Server adds a new tool (most common reconnect drift) → trivial
-  cost, accept silently.
-- Server changes an existing tool's schema or description → bounded
-  cost depending on position, surface a one-line warning.
-- Server completely reorders or replaces the tool list → effectively
-  full miss, refuse or require `--force`.
-
-This nudges the design call away from blanket "strict default"
-toward a **graduated permissive** policy: accept appends silently,
-warn on mid-stream edits, refuse on whole-list reorders or removals.
-
-The strict approach can still be the explicit `--strict` flag for
-users who need every byte of cache (e.g. high-volume scripted runs).

benchmarks/spike-mcp-reconnect/runner.ts

@@ -1,182 +0,0 @@
-/** Empirically confirms RFC #110: tool-list drift mid-session breaks DeepSeek's prefix cache. */
-
-import { DeepSeekClient, loadDotenv } from "../../src/index.js";
-import type { ChatMessage, ToolSpec } from "../../src/types.js";
-
-loadDotenv();
-
-const MODEL = "deepseek-chat";
-
-// DeepSeek's prefix cache only kicks in past ~1024 tokens of repeated
-// prefix, so the system prompt has to be substantial. Padded with
-// realistic-shape filler so the test exercises the same code path a
-// real Reasonix session would.
-const SYSTEM = [
-  "You are a precise senior software engineer assisting with TypeScript codebases.",
-  "Style: terse, concrete, no filler. Don't restate the question. Don't apologise.",
-  "When you reference a file, use the file:line shape so the user can click through.",
-  "When you list options, use a bulleted list, not prose.",
-  "When you reason, do so internally — final reply is the answer, not the deliberation.",
-  "If the user asks a yes/no question, lead with the answer, then one sentence of why.",
-  "If the user asks for code, output a minimal patch — not the whole file.",
-  "When a tool result is a long stack trace, surface the root frame and one stack item up.",
-  "Refuse to fabricate API surface; if you don't know a function exists, ask first.",
-  "Prefer existing libraries over hand-rolled implementations; flag the dependency cost.",
-  "When suggesting a refactor, point at the existing call sites that motivate it.",
-  "Don't add error handling that catches errors you have no recovery for.",
-  "Don't add comments that restate the code; only document the WHY.",
-  "Keep diffs reviewable: 1 change per commit, related changes per PR.",
-  "Treat user messages as authoritative on intent; ask only when truly ambiguous.",
-  "Response length is a function of complexity — short answers when short suffices.",
-  "Default to UTF-8, LF line endings, two-space indent unless the surrounding code differs.",
-  "When a request is impossible as stated, propose the nearest tractable alternative.",
-  "Cite specific paths and line numbers; never paraphrase a file you can read.",
-  "Optimise for the user's next action, not your own throughness.",
-].join(" ");
-
-const baseTool: ToolSpec = {
-  type: "function",
-  function: {
-    name: "read_file",
-    description: "Read the contents of a file at the given absolute or workspace-relative path.",
-    parameters: {
-      type: "object",
-      properties: {
-        path: { type: "string", description: "File path to read." },
-      },
-      required: ["path"],
-    },
-  },
-};
-
-const writeTool: ToolSpec = {
-  type: "function",
-  function: {
-    name: "write_file",
-    description: "Write text content to a file at the given path. Creates parents if missing.",
-    parameters: {
-      type: "object",
-      properties: {
-        path: { type: "string" },
-        content: { type: "string" },
-      },
-      required: ["path", "content"],
-    },
-  },
-};
-
-const searchTool: ToolSpec = {
-  type: "function",
-  function: {
-    name: "search_files",
-    description: "Find files whose path matches a glob pattern under the workspace root.",
-    parameters: {
-      type: "object",
-      properties: { pattern: { type: "string" } },
-      required: ["pattern"],
-    },
-  },
-};
-
-const TOOLSET_A: ToolSpec[] = [baseTool, writeTool, searchTool];
-
-// Same shape as TOOLSET_A but adds one extra tool — emulates an MCP
-// server reconnect that exposed an additional capability.
-const newTool: ToolSpec = {
-  type: "function",
-  function: {
-    name: "delete_file",
-    description: "Remove a file at the given path. Refuses on directories.",
-    parameters: {
-      type: "object",
-      properties: { path: { type: "string" } },
-      required: ["path"],
-    },
-  },
-};
-const TOOLSET_A_PLUS: ToolSpec[] = [...TOOLSET_A, newTool];
-
-// Same set as A, only the description on read_file edited.
-const TOOLSET_A_EDITED: ToolSpec[] = [
-  {
-    ...baseTool,
-    function: {
-      ...baseTool.function,
-      description: "Read the contents of a file at the given path. Returns text or binary base64.",
-    },
-  },
-  writeTool,
-  searchTool,
-];
-
-interface Turn {
-  label: string;
-  tools: ToolSpec[];
-  user: string;
-}
-
-const TURNS: Turn[] = [
-  { label: "1 · cold start (toolset A)        ", tools: TOOLSET_A, user: "What does Reasonix optimise for, in one sentence?" },
-  { label: "2 · same prefix (toolset A)        ", tools: TOOLSET_A, user: "And why DeepSeek-only?" },
-  { label: "3 · drift: ADDED tool (toolset A+) ", tools: TOOLSET_A_PLUS, user: "What's a flaky test, in one sentence?" },
-  { label: "4 · same prefix again (toolset A+) ", tools: TOOLSET_A_PLUS, user: "And how do you stabilise one?" },
-  { label: "5 · drift: EDITED desc (toolset A')", tools: TOOLSET_A_EDITED, user: "What's an idempotent operation, in one sentence?" },
-];
-
-async function main(): Promise<void> {
-  if (!process.env.DEEPSEEK_API_KEY) {
-    console.error("DEEPSEEK_API_KEY not set in env. Add it to .env and re-run.");
-    process.exit(1);
-  }
-  const client = new DeepSeekClient();
-
-  const systemMsg: ChatMessage = { role: "system", content: SYSTEM };
-
-  console.log(`RFC #110 cache spike — DeepSeek prompt-cache behaviour under tool-list drift`);
-  console.log(`model: ${MODEL}`);
-  console.log(`system prompt: ${SYSTEM.length} chars`);
-  console.log();
-  const head =
-    "turn".padEnd(40) +
-    "prompt".padStart(8) +
-    "hit".padStart(8) +
-    "miss".padStart(8) +
-    "hit%".padStart(8) +
-    "ms".padStart(8);
-  console.log(head);
-  console.log("-".repeat(head.length));
-
-  for (const turn of TURNS) {
-    const messages: ChatMessage[] = [systemMsg, { role: "user", content: turn.user }];
-    const t0 = Date.now();
-    const resp = await client.chat({
-      model: MODEL,
-      messages,
-      tools: turn.tools,
-      maxTokens: 64,
-    });
-    const ms = Date.now() - t0;
-    const u = resp.usage;
-    const hitRatio = u.promptTokens > 0 ? (u.promptCacheHitTokens / u.promptTokens) * 100 : 0;
-    console.log(
-      turn.label.padEnd(40) +
-        String(u.promptTokens).padStart(8) +
-        String(u.promptCacheHitTokens).padStart(8) +
-        String(u.promptCacheMissTokens).padStart(8) +
-        `${hitRatio.toFixed(1)}%`.padStart(8) +
-        String(ms).padStart(8),
-    );
-  }
-  console.log();
-  console.log("Expectations:");
-  console.log("  Turn 1 — cold, hit ≈ 0");
-  console.log("  Turn 2 — same prefix as 1, hit ≈ system-prompt-token-count");
-  console.log("  Turn 3 — DRIFTED (added tool), hit drops to ~0 again");
-  console.log("  Turn 4 — same prefix as 3, hit climbs back");
-  console.log("  Turn 5 — DRIFTED (edited description), hit drops to ~0 again");
-}
-
-main().catch((e) => {
-  console.error(e);
-  process.exit(1);
-});

benchmarks/spike-tdd-kernel/bench-latency.mjs

@@ -1,75 +0,0 @@
-import { spawnSync } from "node:child_process";
-import { readFileSync, writeFileSync } from "node:fs";
-
-const samples = [
-  ["tests/checkpoints.test.ts", "snapshots existing files with their content"],
-  ["tests/checkpoints.test.ts", "matches by exact id"],
-  ["tests/compact-tokens.test.ts", "leaves small tool messages alone"],
-  ["tests/compact-tokens.test.ts", "shrinks tool messages that exceed the token budget"],
-  ["tests/diff.test.ts", null],
-  ["tests/edit-blocks.test.ts", null],
-  ["tests/event-replay.test.ts", null],
-  ["tests/event-sink-jsonl.test.ts", null],
-  ["tests/at-mentions.test.ts", null],
-  ["tests/bang.test.ts", null],
-];
-
-function pickFirstIt(file) {
-  const src = readFileSync(file, "utf8");
-  const m = src.match(/^\s*(?:it|test)\(["']([^"']+)["']/m);
-  return m ? m[1] : null;
-}
-
-function runOnce(file, name) {
-  const args = ["vitest", "--run", file];
-  if (name) args.push("-t", name);
-  const t0 = Date.now();
-  const res = spawnSync("npx", args, { encoding: "utf8", shell: true });
-  const ms = Date.now() - t0;
-  return { ms, ok: res.status === 0, stderr: res.stderr.slice(-400) };
-}
-
-function pct(arr, p) {
-  const s = [...arr].sort((a, b) => a - b);
-  return s[Math.floor((s.length - 1) * p)];
-}
-
-const results = [];
-console.log("Sampling test names where missing…");
-for (const s of samples) {
-  if (!s[1]) {
-    s[1] = pickFirstIt(s[0]);
-    console.log(`  ${s[0]} → "${s[1]}"`);
-  }
-}
-
-console.log("\nCold run (each test, first time):");
-const cold = [];
-for (const [file, name] of samples) {
-  const r = runOnce(file, name);
-  cold.push(r.ms);
-  console.log(`  ${file} -t "${name?.slice(0, 40)}…" → ${r.ms}ms  ok=${r.ok}`);
-  results.push({ phase: "cold", file, name, ...r });
-}
-
-console.log("\nWarm run (immediate repeat):");
-const warm = [];
-for (const [file, name] of samples) {
-  const r = runOnce(file, name);
-  warm.push(r.ms);
-  console.log(`  ${file} -t "${name?.slice(0, 40)}…" → ${r.ms}ms  ok=${r.ok}`);
-  results.push({ phase: "warm", file, name, ...r });
-}
-
-const summary = {
-  cold: { median: pct(cold, 0.5), p95: pct(cold, 0.95), max: Math.max(...cold) },
-  warm: { median: pct(warm, 0.5), p95: pct(warm, 0.95), max: Math.max(...warm) },
-};
-
-console.log("\n=== Summary ===");
-console.log(JSON.stringify(summary, null, 2));
-
-writeFileSync(
-  new URL("./latency.json", import.meta.url),
-  JSON.stringify({ summary, runs: results }, null, 2),
-);

benchmarks/spike-tdd-kernel/cost-results.json

@@ -1,106 +0,0 @@
-{
-  "summary": {
-    "A_baseline": {
-      "warm": 0,
-      "hot": 0.8347826086956521,
-      "hot2": 0.8347826086956521
-    },
-    "B_augmented": {
-      "warm": 0.6969147005444646,
-      "hot": 0.9360146252285192,
-      "hot2": 0.9360146252285192
-    },
-    "delta_hot": 0.10123201653286706,
-    "delta_hot2": 0.10123201653286706,
-    "pass_A_hot": false,
-    "pass_B_hot": true
-  },
-  "A": {
-    "warm": {
-      "ms": 835,
-      "usage": {
-        "prompt_tokens": 464,
-        "completion_tokens": 1,
-        "total_tokens": 465,
-        "prompt_tokens_details": {
-          "cached_tokens": 0
-        },
-        "prompt_cache_hit_tokens": 0,
-        "prompt_cache_miss_tokens": 464
-      },
-      "ratio": 0
-    },
-    "hot": {
-      "ms": 1901,
-      "usage": {
-        "prompt_tokens": 460,
-        "completion_tokens": 120,
-        "total_tokens": 580,
-        "prompt_tokens_details": {
-          "cached_tokens": 384
-        },
-        "prompt_cache_hit_tokens": 384,
-        "prompt_cache_miss_tokens": 76
-      },
-      "ratio": 0.8347826086956521
-    },
-    "hot2": {
-      "ms": 2792,
-      "usage": {
-        "prompt_tokens": 460,
-        "completion_tokens": 200,
-        "total_tokens": 660,
-        "prompt_tokens_details": {
-          "cached_tokens": 384
-        },
-        "prompt_cache_hit_tokens": 384,
-        "prompt_cache_miss_tokens": 76
-      },
-      "ratio": 0.8347826086956521
-    }
-  },
-  "B": {
-    "warm": {
-      "ms": 575,
-      "usage": {
-        "prompt_tokens": 551,
-        "completion_tokens": 2,
-        "total_tokens": 553,
-        "prompt_tokens_details": {
-          "cached_tokens": 384
-        },
-        "prompt_cache_hit_tokens": 384,
-        "prompt_cache_miss_tokens": 167
-      },
-      "ratio": 0.6969147005444646
-    },
-    "hot": {
-      "ms": 2065,
-      "usage": {
-        "prompt_tokens": 547,
-        "completion_tokens": 120,
-        "total_tokens": 667,
-        "prompt_tokens_details": {
-          "cached_tokens": 512
-        },
-        "prompt_cache_hit_tokens": 512,
-        "prompt_cache_miss_tokens": 35
-      },
-      "ratio": 0.9360146252285192
-    },
-    "hot2": {
-      "ms": 1959,
-      "usage": {
-        "prompt_tokens": 547,
-        "completion_tokens": 120,
-        "total_tokens": 667,
-        "prompt_tokens_details": {
-          "cached_tokens": 512
-        },
-        "prompt_cache_hit_tokens": 512,
-        "prompt_cache_miss_tokens": 35
-      },
-      "ratio": 0.9360146252285192
-    }
-  }
-}

benchmarks/spike-tdd-kernel/cost-results.md

@@ -1,77 +0,0 @@
-# Exp 1 — cache-hit cost analysis
-
-**Result: PASS.** Augmenting `edit_file` tool_results with an `[edit_claim]` + `[test_run]` footer does **not** reduce cache hit. In a controlled side-by-side, the augmented variant cache-hit at **93.6%** vs the baseline's **83.5%** on the same hot turn — a **+10pt improvement**, not a regression.
-
-This makes sense once you reason about where the new tokens land: they sit *inside the prefix*, not *in the tail*. On every subsequent turn they cache-hit. The non-cacheable tail (the new user message) is the same size in both variants, so growing the prefix grows the cache-hit ratio.
-
-## Method
-
-`benchmarks/spike-tdd-kernel/cost.mjs`. Two synthetic 4-turn agent transcripts, identical except that variant B's `edit_file` tool_result carries the RFC's proposed footer:
-
-```
-[test_run] test_id="…" status="pass" duration_ms=1873 command="npx vitest …"
-[edit_claim] test_id="…" edit_target="src/util/slugify.ts" satisfied=true
-```
-
-For each variant, three calls in sequence on `deepseek-chat`:
-1. **warmup** — seeds the prefix into DeepSeek's cache.
-2. **hot** — same prefix + a different small tail, measures steady-state cache hit.
-3. **hot2** — repeat to confirm stability.
-
-Cache hit ratio = `prompt_cache_hit_tokens / (hit + miss)` from the `usage` object.
-
-Raw runs in `cost-results.json`.
-
-## Numbers
-
-```
-                     prompt   hit   miss   ratio   wall
-A_baseline.warmup     464      0    464    0.0%    835ms
-A_baseline.hot        460    384     76   83.5%   1901ms
-A_baseline.hot2       460    384     76   83.5%   2792ms
-
-B_augmented.warmup    551    384    167   69.7%    575ms
-B_augmented.hot       547    512     35   93.6%   2065ms
-B_augmented.hot2      547    512     35   93.6%   1959ms
-```
-
-`B_augmented.warmup` already shows 69.7% because A's system prompt is in cache from prior calls — same byte-stable prefix region.
-
-## Why B has a *better* ratio than A
-
-The augmentation adds ~87 tokens to the prefix (the `[edit_claim]`/`[test_run]` footer). On the hot turn:
-
-- A: prefix-cacheable = 384 tok, tail = 76 tok → 384 / (384+76) = 83.5%
-- B: prefix-cacheable = 512 tok, tail = 35 tok → 512 / (512+35) = 93.6%
-
-Both have the same kind of tail (a new user message). B's tail is smaller because the model emitted a slightly different response continuation seed; nonetheless, the structural point holds: **augmenting tool_results moves bytes from "uncached" (this-turn-only) to "cached" (re-used by every subsequent turn)**.
-
-In real Reasonix sessions with multi-thousand-token histories, the absolute cache-hit ratio is dominated by history size; the marginal effect of an extra ~80 tokens per edit is to *raise* it slightly, not lower it.
-
-## Pass criterion (revised)
-
-The original RFC threshold of "≥92% absolute" doesn't apply cleanly to this synthetic harness — the transcript is only ~460 tokens, far smaller than a typical Reasonix session, which inflates the tail's relative weight.
-
-The substantive criterion is **no degradation**:
-
-> augmentation must not reduce cache hit by more than 2pts vs baseline
-
-Observed: **+10pt improvement**. Passes trivially.
-
-## Implications for the RFC
-
-1. **Cost story is intact.** The "kept cache hit ≥94%" claim in the README is unaffected. Augmenting tool_results is cache-positive, not cache-negative.
-
-2. **Footer placement matters.** Two safe places:
-   - **Append to `edit_file` tool_result** (this experiment). Cache-friendly.
-   - **Insert as a separate synthetic `tool` message between turns** (would also be cache-friendly *if* always at the same position).
-
-   Avoid: rewriting an old tool_result mid-stream, which would invalidate cache from that point onward. The `AppendOnlyLog` invariant in `src/loop.ts` already prevents this.
-
-3. **Footer format should be deterministic.** No timestamps that change per cache-hit attempt; no run-relative durations that vary; no random IDs. The fields chosen (`test_id`, `status`, `duration_ms`, `command`) are all deterministic at write time and frozen thereafter — same bytes, same cache.
-
-4. **Token cost is real but small.** ~80 prompt tokens per edit on subsequent turns. At v4-flash pricing that's negligible. The model also uses ~20 completion tokens to emit `edit_claim`. Total marginal cost per edit: <$0.0001.
-
-## Decision
-
-Greenlight Exp 1. **All four spike experiments pass.** Ready to comment "spike green" on #25 and start a 48h FCP.

benchmarks/spike-tdd-kernel/cost.mjs

@@ -1,154 +0,0 @@
-// Exp 1 — cost: does augmenting tool_result with test_run footers drop cache hit?
-//
-// Approach: build two synthetic 4-turn agent transcripts, identical except that
-// variant B's tool_results carry an extra "[test_run: …]" footer. For each
-// variant, send a "warmup" call to seed the prefix cache, then a "hot" call
-// with a small tail change. Measure cache hit ratio on the hot call.
-//
-// Hypothesis: ratios within ±2 pts; both ≥92%.
-
-import { readFileSync, writeFileSync } from "node:fs";
-
-for (const line of readFileSync(new URL("../../.env", import.meta.url), "utf8").split(/\r?\n/)) {
-  const m = line.match(/^([A-Z_]+)\s*=\s*(.*)$/);
-  if (m && !process.env[m[1]]) process.env[m[1]] = m[2].replace(/^["']|["']$/g, "");
-}
-const apiKey = process.env.DEEPSEEK_API_KEY;
-const baseUrl = process.env.DEEPSEEK_BASE_URL ?? "https://api.deepseek.com/v1";
-
-const SYS = `You are a coding agent working in a TypeScript repository.
-You have these tools available: read_file, edit_file, run_command, list_dir.
-Reply concisely. Do NOT call any tool in this synthetic transcript — just respond in plain text when asked.`;
-
-// 4-turn synthetic transcript with three tool_result messages.
-const transcript = [
-  { role: "system", content: SYS },
-  { role: "user", content: "Find where the slugify helper is defined and show me its current implementation." },
-  {
-    role: "assistant",
-    content: "",
-    tool_calls: [
-      { id: "c1", type: "function", function: { name: "list_dir", arguments: JSON.stringify({ path: "src/util" }) } },
-    ],
-  },
-  {
-    role: "tool",
-    tool_call_id: "c1",
-    content: "src/util/clamp.ts\nsrc/util/color.ts\nsrc/util/duration.ts\nsrc/util/ranges.ts\nsrc/util/ring.ts\nsrc/util/slugify.ts\nsrc/util/uniq.ts",
-  },
-  {
-    role: "assistant",
-    content: "",
-    tool_calls: [
-      { id: "c2", type: "function", function: { name: "read_file", arguments: JSON.stringify({ path: "src/util/slugify.ts" }) } },
-    ],
-  },
-  {
-    role: "tool",
-    tool_call_id: "c2",
-    content: `export function slugify(s: string): string {
-  return s
-    .toLowerCase()
-    .replace(/[^a-z0-9]+/g, "-")
-    .replace(/^-+|-+$/g, "");
-}`,
-  },
-  { role: "user", content: "Now narrow it: collapse repeated dashes too." },
-  {
-    role: "assistant",
-    content: "",
-    tool_calls: [
-      { id: "c3", type: "function", function: { name: "edit_file", arguments: JSON.stringify({ path: "src/util/slugify.ts", search: ".replace(/[^a-z0-9]+/g, \"-\")", replace: ".replace(/[^a-z0-9]+/g, \"-\").replace(/-{2,}/g, \"-\")" }) } },
-    ],
-  },
-  // baseline tool_result
-  {
-    role: "tool",
-    tool_call_id: "c3",
-    content: "Edit applied: src/util/slugify.ts (1 hunk).",
-  },
-  { role: "user", content: "Good, now also trim leading/trailing whitespace before the regex pass." },
-];
-
-// Variant B: same transcript, but the edit_file tool_result also carries a test_run footer.
-// This is the EXACT extra payload the RFC would inject.
-const augmentedToolResult = `Edit applied: src/util/slugify.ts (1 hunk).
-
-[test_run] test_id="tests/slugify.test.ts::slugify collapses repeated dashes" status="pass" duration_ms=1873 command="npx vitest --run tests/slugify.test.ts -t \\"collapses repeated dashes\\""
-[edit_claim] test_id="tests/slugify.test.ts::slugify collapses repeated dashes" edit_target="src/util/slugify.ts" satisfied=true`;
-
-function variantA() {
-  return JSON.parse(JSON.stringify(transcript));
-}
-
-function variantB() {
-  const t = JSON.parse(JSON.stringify(transcript));
-  // augment the edit_file tool result (index 8)
-  t[8].content = augmentedToolResult;
-  return t;
-}
-
-async function call(messages, tag) {
-  const t0 = Date.now();
-  const resp = await fetch(`${baseUrl}/chat/completions`, {
-    method: "POST",
-    headers: { Authorization: `Bearer ${apiKey}`, "Content-Type": "application/json" },
-    body: JSON.stringify({
-      model: "deepseek-chat",
-      messages,
-      temperature: 0,
-      max_tokens: 200,
-      stream: false,
-      // Thinking off so synthetic assistant messages don't need reasoning_content round-trip.
-      // Cache mechanic is byte-prefix; thinking on/off doesn't change that.
-      extra_body: { thinking: { type: "disabled" } },
-    }),
-  });
-  if (!resp.ok) throw new Error(`API ${resp.status}: ${await resp.text()}`);
-  const data = await resp.json();
-  const ms = Date.now() - t0;
-  const u = data.usage;
-  const hit = u.prompt_cache_hit_tokens ?? 0;
-  const miss = u.prompt_cache_miss_tokens ?? 0;
-  const ratio = hit + miss > 0 ? hit / (hit + miss) : 0;
-  console.log(
-    `  [${tag}] prompt=${u.prompt_tokens} hit=${hit} miss=${miss} ratio=${(ratio * 100).toFixed(1)}%  ${ms}ms`,
-  );
-  return { ms, usage: u, ratio };
-}
-
-async function runVariant(name, build) {
-  console.log(`\n--- variant ${name} ---`);
-  const warmupTail = { role: "user", content: "(warmup ping — just say 'ok')" };
-  const realTail = { role: "user", content: "Show me the resulting file." };
-
-  // 1. warmup — seed the cache
-  const warm = await call([...build(), warmupTail], `${name}.warmup`);
-  // 2. hot — same prefix, different tail
-  const hot = await call([...build(), realTail], `${name}.hot`);
-  // 3. hot-2 — repeat to confirm cache stickiness
-  const hot2 = await call([...build(), realTail], `${name}.hot2`);
-
-  return { warm, hot, hot2 };
-}
-
-console.log("Exp 1 — cache hit comparison");
-const A = await runVariant("A_baseline", variantA);
-const B = await runVariant("B_augmented", variantB);
-
-const summary = {
-  A_baseline: { warm: A.warm.ratio, hot: A.hot.ratio, hot2: A.hot2.ratio },
-  B_augmented: { warm: B.warm.ratio, hot: B.hot.ratio, hot2: B.hot2.ratio },
-  delta_hot: B.hot.ratio - A.hot.ratio,
-  delta_hot2: B.hot2.ratio - A.hot2.ratio,
-  pass_A_hot: A.hot.ratio >= 0.92,
-  pass_B_hot: B.hot.ratio >= 0.92,
-};
-
-console.log("\n=== Summary ===");
-console.log(JSON.stringify(summary, null, 2));
-
-writeFileSync(
-  new URL("./cost-results.json", import.meta.url),
-  JSON.stringify({ summary, A, B }, null, 2),
-);

benchmarks/spike-tdd-kernel/latency.json

@@ -1,176 +0,0 @@
-{
-  "summary": {
-    "cold": {
-      "median": 1900,
-      "p95": 4731,
-      "max": 4815
-    },
-    "warm": {
-      "median": 1888,
-      "p95": 4972,
-      "max": 5075
-    }
-  },
-  "runs": [
-    {
-      "phase": "cold",
-      "file": "tests/checkpoints.test.ts",
-      "name": "snapshots existing files with their content",
-      "ms": 1705,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/checkpoints.test.ts",
-      "name": "matches by exact id",
-      "ms": 1584,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/compact-tokens.test.ts",
-      "name": "leaves small tool messages alone",
-      "ms": 2130,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/compact-tokens.test.ts",
-      "name": "shrinks tool messages that exceed the token budget",
-      "ms": 2362,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/diff.test.ts",
-      "name": "returns 1 for identical strings",
-      "ms": 1900,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/edit-blocks.test.ts",
-      "name": "parses a single block",
-      "ms": 4731,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/event-replay.test.ts",
-      "name": "synthetic LoopEvents → eventize → sink → file → source → reducers → ConversationView matches",
-      "ms": 1668,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/event-sink-jsonl.test.ts",
-      "name": "appends one JSON object per line, parseable round-trip",
-      "ms": 2574,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/at-mentions.test.ts",
-      "name": "matches @path at start of string",
-      "ms": 1897,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "cold",
-      "file": "tests/bang.test.ts",
-      "name": "returns the command body for a `!`-prefixed input",
-      "ms": 4815,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/checkpoints.test.ts",
-      "name": "snapshots existing files with their content",
-      "ms": 1626,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/checkpoints.test.ts",
-      "name": "matches by exact id",
-      "ms": 1585,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/compact-tokens.test.ts",
-      "name": "leaves small tool messages alone",
-      "ms": 2027,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/compact-tokens.test.ts",
-      "name": "shrinks tool messages that exceed the token budget",
-      "ms": 2301,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/diff.test.ts",
-      "name": "returns 1 for identical strings",
-      "ms": 1888,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/edit-blocks.test.ts",
-      "name": "parses a single block",
-      "ms": 4972,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/event-replay.test.ts",
-      "name": "synthetic LoopEvents → eventize → sink → file → source → reducers → ConversationView matches",
-      "ms": 1767,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/event-sink-jsonl.test.ts",
-      "name": "appends one JSON object per line, parseable round-trip",
-      "ms": 2523,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/at-mentions.test.ts",
-      "name": "matches @path at start of string",
-      "ms": 1883,
-      "ok": true,
-      "stderr": ""
-    },
-    {
-      "phase": "warm",
-      "file": "tests/bang.test.ts",
-      "name": "returns the command body for a `!`-prefixed input",
-      "ms": 5075,
-      "ok": true,
-      "stderr": ""
-    }
-  ]
-}

benchmarks/spike-tdd-kernel/latency.md

@@ -1,36 +0,0 @@
-# Exp 4 — `vitest -t` latency on this repo
-
-**Result: PASS.** Median 1.9s, p95 ~5.0s, max 5.1s. Both pass thresholds met (median ≤3s, p95 ≤6s).
-
-## Method
-
-`benchmarks/spike-tdd-kernel/bench-latency.mjs` runs `npx vitest --run <file> -t "<name>"` against 10 sampled test/name pairs across 9 different test files, twice each (cold = first invocation, warm = immediate repeat). Each invocation is a fresh `npx` subprocess. Wall-clock measured around `spawnSync`. Raw data in `latency.json`.
-
-## Numbers
-
-| | median | p95 | max |
-|---|---|---|---|
-| cold | 1900 ms | 4731 ms | 4815 ms |
-| warm | 1888 ms | 4972 ms | 5075 ms |
-
-All 20 invocations exited 0.
-
-## Findings
-
-1. **Cold ≈ warm.** Each `npx vitest --run` boots a fresh worker, so there is no meaningful warm-up benefit. The ~1.9s floor is overwhelmingly framework startup (vite + vitest + tsx transform), not test work. The two slowest tests (`edit-blocks`, `bang`) hit ~5s on both cold and warm, indicating per-test overhead specifically — likely module graph size, not test logic.
-
-2. **Implication for kernel design.** Running N separate `vitest --run -t <id_n>` is N × ~2s. **Batching multiple `test_id`s in one invocation** (`vitest --run -t a -t b -t c`) almost certainly amortises the boot cost. RFC's "auto-run after each edit" should bundle test_ids when an edit pass writes more than one — and a bulk-edit batch should only fire one vitest invocation at the end.
-
-3. **Threshold headroom is thin on slow tests.** A test that already takes 5s warm leaves ~1s for kernel overhead before the user starts noticing. Per-edit auto-run is fine; per-keystroke would not be.
-
-## Decision
-
-Greenlight the latency assumption in the RFC. Update RFC §"Cost analysis" to reflect:
-- "+1 test run per edit" → "+1 vitest invocation per edit batch"
-- Add note that the kernel should coalesce edits within one model turn into a single `vitest -t a -t b …` call.
-
-## Sample tests used
-
-- `checkpoints.test.ts` (×2)
-- `compact-tokens.test.ts` (×2)
-- `diff.test.ts`, `edit-blocks.test.ts`, `event-replay.test.ts`, `event-sink-jsonl.test.ts`, `at-mentions.test.ts`, `bang.test.ts`

benchmarks/spike-tdd-kernel/tdd-eval.json

@@ -1,387 +0,0 @@
-{
-  "passed": 8,
-  "total": 10,
-  "totalUsage": {
-    "prompt_tokens": 2246,
-    "completion_tokens": 5732
-  },
-  "runs": [
-    {
-      "id": "e1",
-      "level": "easy",
-      "task": "A pure function `slugify(s: string): string` in src/util/slugify.ts that lowercases, replaces non-alphanumerics with '-', and collapses repeated dashes.",
-      "ms": 9299,
-      "usage": {
-        "prompt_tokens": 207,
-        "completion_tokens": 707,
-        "total_tokens": 914,
-        "prompt_tokens_details": {
-          "cached_tokens": 0
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 454
-        },
-        "prompt_cache_hit_tokens": 0,
-        "prompt_cache_miss_tokens": 207
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": false,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": false,
-        "target": "src/util/slugify",
-        "symbol": "slugify",
-        "itNames": [
-          "should lowercase the input",
-          "should replace non-alphanumeric characters with hyphens",
-          "should collapse consecutive hyphens into one",
-          "should trim leading and trailing hyphens",
-          "should handle an empty string",
-          "should handle a string with only non-alphanumeric characters",
-          "should preserve alphanumeric characters and hyphens from original"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { slugify } from '../slugify';\n\ndescribe('slugify', () => {\n  it('should lowercase the input', () => {\n    expect(slugify('HELLO')).toBe('hello');\n  });\n\n  it('should replace non-alphanumeric characters with hyphens', () => {\n    expect(slugify('hello world!')).toBe('hello-world');\n  });\n\n  it('should collapse consecutive hyphens into one', () => {\n    expect(slugify('hello---world')).toBe('hello-world');\n  });\n\n  it('should trim leading and trailing hyphens', () => {\n    expect(slugify('--hello world--')).toBe('hello-world');\n  });\n\n  it('should handle an empty string', () => {\n    expect(slugify('')).toBe('');\n  });\n\n  it('should handle a string with only non-alphanumeric characters', () => {\n    expect(slugify('!!@#$%^&*()')).toBe('');\n  });\n\n  it('should preserve alphanumeric characters and hyphens from original', () => {\n    expect(slugify('a-B_c')).toBe('a-b_c');\n  });\n});"
-      }
-    },
-    {
-      "id": "e2",
-      "level": "easy",
-      "task": "A pure function `clamp(n: number, lo: number, hi: number): number` in src/util/clamp.ts that clamps n into [lo, hi].",
-      "ms": 4321,
-      "usage": {
-        "prompt_tokens": 205,
-        "completion_tokens": 309,
-        "total_tokens": 514,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 109
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 77
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/clamp",
-        "symbol": "clamp",
-        "itNames": [
-          "should return the lower bound when n is below the range",
-          "should return the upper bound when n is above the range",
-          "should return the original number when n is within the range",
-          "should handle floating point numbers correctly",
-          "should work when lo equals hi"
-        ],
-        "code": "import { clamp } from '../src/util/clamp';\n\ndescribe('clamp', () => {\n  it('should return the lower bound when n is below the range', () => {\n    expect(clamp(-5, 0, 10)).toBe(0);\n  });\n\n  it('should return the upper bound when n is above the range', () => {\n    expect(clamp(15, 0, 10)).toBe(10);\n  });\n\n  it('should return the original number when n is within the range', () => {\n    expect(clamp(5, 0, 10)).toBe(5);\n  });\n\n  it('should handle floating point numbers correctly', () => {\n    expect(clamp(3.14, 0, 5)).toBe(3.14);\n  });\n\n  it('should work when lo equals hi', () => {\n    expect(clamp(100, 5, 5)).toBe(5);\n  });\n});"
-      }
-    },
-    {
-      "id": "e3",
-      "level": "easy",
-      "task": "A pure function `hexToRgb(hex: string): {r:number,g:number,b:number} | null` in src/util/color.ts. Accepts '#abc', '#aabbcc', and 'aabbcc'. Returns null on invalid.",
-      "ms": 8012,
-      "usage": {
-        "prompt_tokens": 224,
-        "completion_tokens": 629,
-        "total_tokens": 853,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 326
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 96
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/color",
-        "symbol": "hexToRgb",
-        "itNames": [
-          "should convert shorthand hex ",
-          "should convert full hex ",
-          "should convert hex without hash ",
-          "should return null for an invalid hex string"
-        ],
-        "code": "import { hexToRgb } from '../src/util/color'\nimport { describe, it, expect } from 'vitest'\n\ndescribe('hexToRgb', () => {\n  it('should convert shorthand hex \"#abc\" to { r: 170, g: 187, b: 204 }', () => {\n    expect(hexToRgb('#abc')).toEqual({ r: 170, g: 187, b: 204 })\n  })\n\n  it('should convert full hex \"#aabbcc\" to { r: 170, g: 187, b: 204 }', () => {\n    expect(hexToRgb('#aabbcc')).toEqual({ r: 170, g: 187, b: 204 })\n  })\n\n  it('should convert hex without hash \"aabbcc\" to { r: 170, g: 187, b: 204 }', () => {\n    expect(hexToRgb('aabbcc')).toEqual({ r: 170, g: 187, b: 204 })\n  })\n\n  it('should return null for an invalid hex string', () => {\n    expect(hexToRgb('xyz')).toBeNull()\n    expect(hexToRgb('#ggg')).toBeNull()\n    expect(hexToRgb('#12345')).toBeNull()\n    expect(hexToRgb('')).toBeNull()\n  })\n})"
-      }
-    },
-    {
-      "id": "e4",
-      "level": "easy",
-      "task": "A pure function `uniqueBy<T,K>(arr: T[], key: (t: T) => K): T[]` in src/util/uniq.ts preserving first occurrence.",
-      "ms": 9055,
-      "usage": {
-        "prompt_tokens": 206,
-        "completion_tokens": 687,
-        "total_tokens": 893,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 328
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 78
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/uniq",
-        "symbol": "uniqueBy",
-        "itNames": [
-          "should return array with unique elements based on key function, preserving first occurrence",
-          "should handle empty array",
-          "should work with primitive keys and preserve order",
-          "should handle objects with different key types"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { uniqueBy } from '../src/util/uniq';\n\ndescribe('uniqueBy', () => {\n  it('should return array with unique elements based on key function, preserving first occurrence', () => {\n    const input = [{ id: 1, name: 'a' }, { id: 2, name: 'b' }, { id: 1, name: 'c' }];\n    const result = uniqueBy(input, (item) => item.id);\n    expect(result).toEqual([{ id: 1, name: 'a' }, { id: 2, name: 'b' }]);\n  });\n\n  it('should handle empty array', () => {\n    expect(uniqueBy([], (x: unknown) => x)).toEqual([]);\n  });\n\n  it('should work with primitive keys and preserve order', () => {\n    const arr = [3, 1, 2, 3, 4, 1, 5];\n    const result = uniqueBy(arr, (x) => x);\n    expect(result).toEqual([3, 1, 2, 4, 5]);\n  });\n\n  it('should handle objects with different key types', () => {\n    const items = [\n      { val: 'x', group: 1 },\n      { val: 'y', group: 2 },\n      { val: 'z', group: 1 },\n    ];\n    const result = uniqueBy(items, (item) => item.group);\n    expect(result).toEqual([\n      { val: 'x', group: 1 },\n      { val: 'y', group: 2 },\n    ]);\n  });\n});"
-      }
-    },
-    {
-      "id": "e5",
-      "level": "easy",
-      "task": "A pure function `parseDuration(s: string): number` in src/util/duration.ts. '1500ms' → 1500, '2s' → 2000, '1m' → 60000. Returns NaN on invalid.",
-      "ms": 3168,
-      "usage": {
-        "prompt_tokens": 221,
-        "completion_tokens": 234,
-        "total_tokens": 455,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 67
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 93
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/duration",
-        "symbol": "parseDuration",
-        "itNames": [
-          "should return 1500 for ",
-          "should return 2000 for ",
-          "should return 60000 for ",
-          "should return NaN for invalid input"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { parseDuration } from '../src/util/duration';\n\ndescribe('parseDuration', () => {\n  it('should return 1500 for \"1500ms\"', () => {\n    expect(parseDuration('1500ms')).toBe(1500);\n  });\n\n  it('should return 2000 for \"2s\"', () => {\n    expect(parseDuration('2s')).toBe(2000);\n  });\n\n  it('should return 60000 for \"1m\"', () => {\n    expect(parseDuration('1m')).toBe(60000);\n  });\n\n  it('should return NaN for invalid input', () => {\n    expect(parseDuration('invalid')).toBeNaN();\n  });\n});"
-      }
-    },
-    {
-      "id": "m1",
-      "level": "medium",
-      "task": "A class `RingBuffer<T>` in src/util/ring.ts with capacity, push(x) (drops oldest when full), toArray() returning oldest-first, and size getter.",
-      "ms": 3326,
-      "usage": {
-        "prompt_tokens": 207,
-        "completion_tokens": 237,
-        "total_tokens": 444,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 112
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 79
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/ring",
-        "symbol": "RingBuffer",
-        "itNames": [
-          "should drop oldest element when full and maintain capacity"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { RingBuffer } from '../src/util/ring';\n\ndescribe('RingBuffer', () => {\n  it('should drop oldest element when full and maintain capacity', () => {\n    const buffer = new RingBuffer<number>(3);\n    buffer.push(1);\n    buffer.push(2);\n    buffer.push(3);\n    buffer.push(4); // should drop 1\n    expect(buffer.size).toBe(3);\n    expect(buffer.toArray()).toEqual([2, 3, 4]);\n  });\n});"
-      }
-    },
-    {
-      "id": "m2",
-      "level": "medium",
-      "task": "A function `mergeRanges(ranges: Array<[number,number]>): Array<[number,number]>` in src/util/ranges.ts. Coalesces overlapping/adjacent ranges, returns sorted.",
-      "ms": 5573,
-      "usage": {
-        "prompt_tokens": 213,
-        "completion_tokens": 422,
-        "total_tokens": 635,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 162
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 85
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/ranges",
-        "symbol": "mergeRanges",
-        "itNames": [
-          "should merge overlapping ranges",
-          "should merge adjacent ranges",
-          "should return empty array for empty input",
-          "should handle single range"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { mergeRanges } from '../src/util/ranges';\n\ndescribe('mergeRanges', () => {\n  it('should merge overlapping ranges', () => {\n    const input: Array<[number, number]> = [[1, 3], [2, 6], [8, 10], [15, 18]];\n    const expected: Array<[number, number]> = [[1, 6], [8, 10], [15, 18]];\n    expect(mergeRanges(input)).toEqual(expected);\n  });\n\n  it('should merge adjacent ranges', () => {\n    const input: Array<[number, number]> = [[1, 2], [2, 3], [3, 4]];\n    const expected: Array<[number, number]> = [[1, 4]];\n    expect(mergeRanges(input)).toEqual(expected);\n  });\n\n  it('should return empty array for empty input', () => {\n    expect(mergeRanges([])).toEqual([]);\n  });\n\n  it('should handle single range', () => {\n    expect(mergeRanges([[5, 10]])).toEqual([[5, 10]]);\n  });\n});"
-      }
-    },
-    {
-      "id": "m3",
-      "level": "medium",
-      "task": "A function `debounceAsync<T extends any[], R>(fn: (...args: T) => Promise<R>, ms: number): (...args: T) => Promise<R>` in src/util/debounce.ts. Resolves only the latest call's promise; earlier callers reject with an AbortError-like.",
-      "ms": 7104,
-      "usage": {
-        "prompt_tokens": 233,
-        "completion_tokens": 516,
-        "total_tokens": 749,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 267
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 105
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/util/debounce",
-        "symbol": "debounceAsync",
-        "itNames": [
-          "should resolve the latest call and reject previous calls",
-          "should only call the underlying function once per burst"
-        ],
-        "code": "import { describe, it, expect, vi } from 'vitest';\nimport { debounceAsync } from '../src/util/debounce';\n\ndescribe('debounceAsync', () => {\n  it('should resolve the latest call and reject previous calls', async () => {\n    const fn = vi.fn().mockImplementation(async (x: number) => x * 2);\n    const debounced = debounceAsync(fn, 50);\n\n    const promise1 = debounced(1);\n    const promise2 = debounced(2);\n    const promise3 = debounced(3);\n\n    await expect(promise1).rejects.toThrow();\n    await expect(promise2).rejects.toThrow();\n    await expect(promise3).resolves.toBe(6);\n  });\n\n  it('should only call the underlying function once per burst', async () => {\n    const fn = vi.fn().mockResolvedValue('resolved');\n    const debounced = debounceAsync(fn, 50);\n\n    debounced('a');\n    debounced('b');\n    await debounced('c');\n\n    expect(fn).toHaveBeenCalledTimes(1);\n  });\n});"
-      }
-    },
-    {
-      "id": "h1",
-      "level": "hard",
-      "task": "A function `extractTestId(file: string, fullName: string, source: string): { id: string, source: 'native' | 'annotation' }` in src/repair/test-id.ts. If `source` contains a '// @reasonix-test-id: <slug>' comment within 3 lines above an it()/test() whose name matches `fullName`, return that slug with source='annotation'. Otherwise return `${file}::${fullName}` with source='native'.",
-      "ms": 14142,
-      "usage": {
-        "prompt_tokens": 273,
-        "completion_tokens": 1129,
-        "total_tokens": 1402,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 327
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 145
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": false,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": false,
-        "target": "src/repair/test-id",
-        "symbol": "extractTestId",
-        "itNames": [
-          "returns native id when no annotation comment exists",
-          "should do something",
-          "returns annotation id when matching comment is within 3 lines above",
-          "works",
-          "respects the 3-line limit: exactly 3 lines above matches",
-          "three lines",
-          "does not match comment more than 3 lines above (4 lines)",
-          "too far",
-          "does not match comment for a different test name",
-          "different test",
-          "picks correct comment when multiple exist",
-          "first test",
-          "second test",
-          "does not misidentify a non-comment line with @reasonix-test-id",
-          "no comment"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { extractTestId } from './src/repair/test-id';\n\ndescribe('extractTestId', () => {\n  it('returns native id when no annotation comment exists', () => {\n    const file = 'a.test.ts';\n    const fullName = 'should do something';\n    const source = `\n      it('should do something', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'a.test.ts::should do something', source: 'native' });\n  });\n\n  it('returns annotation id when matching comment is within 3 lines above', () => {\n    const file = 'b.test.ts';\n    const fullName = 'works';\n    const source = `\n      // @reasonix-test-id: my-slug\n      it('works', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'my-slug', source: 'annotation' });\n  });\n\n  it('respects the 3-line limit: exactly 3 lines above matches', () => {\n    const file = 'c.test.ts';\n    const fullName = 'three lines';\n    const source = `\n      // @reasonix-test-id: three-slug\n      // line1\n      // line2\n      it('three lines', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'three-slug', source: 'annotation' });\n  });\n\n  it('does not match comment more than 3 lines above (4 lines)', () => {\n    const file = 'd.test.ts';\n    const fullName = 'too far';\n    const source = `\n      // @reasonix-test-id: wrong\n      // line1\n      // line2\n      // line3\n      it('too far', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'd.test.ts::too far', source: 'native' });\n  });\n\n  it('does not match comment for a different test name', () => {\n    const file = 'e.test.ts';\n    const fullName = 'other test';\n    const source = `\n      // @reasonix-test-id: not-for-you\n      it('different test', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'e.test.ts::other test', source: 'native' });\n  });\n\n  it('picks correct comment when multiple exist', () => {\n    const file = 'f.test.ts';\n    const fullName = 'second test';\n    const source = `\n      // @reasonix-test-id: first-slug\n      it('first test', () => {});\n      // @reasonix-test-id: second-slug\n      it('second test', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'second-slug', source: 'annotation' });\n  });\n\n  it('does not misidentify a non-comment line with @reasonix-test-id', () => {\n    const file = 'g.test.ts';\n    const fullName = 'no comment';\n    const source = `\n      const x = '// @reasonix-test-id: ignored';\n      it('no comment', () => {});\n    `;\n    const result = extractTestId(file, fullName, source);\n    expect(result).toEqual({ id: 'g.test.ts::no comment', source: 'native' });\n  });\n});"
-      }
-    },
-    {
-      "id": "h2",
-      "level": "hard",
-      "task": "A function `pairRedGreen(events: Array<{type:string, test_id?:string, status?:string, ts:number}>): Array<{ test_id: string, red_ts: number, green_ts: number }>` in src/events/pair.ts. For each test_id, find the most recent fail→pass transition and return one entry per test_id. Ignore test_ids that never went green.",
-      "ms": 11174,
-      "usage": {
-        "prompt_tokens": 257,
-        "completion_tokens": 862,
-        "total_tokens": 1119,
-        "prompt_tokens_details": {
-          "cached_tokens": 128
-        },
-        "completion_tokens_details": {
-          "reasoning_tokens": 480
-        },
-        "prompt_cache_hit_tokens": 128,
-        "prompt_cache_miss_tokens": 129
-      },
-      "score": {
-        "compiles_shape": true,
-        "importsTarget": true,
-        "implLeak": false,
-        "stableNames": true,
-        "tsOk": true,
-        "tsErr": "",
-        "passAll": true,
-        "target": "src/events/pair",
-        "symbol": "pairRedGreen",
-        "itNames": [
-          "should return the most recent fail→pass transition per test_id, ignoring test_ids that never passed"
-        ],
-        "code": "import { describe, it, expect } from 'vitest';\nimport { pairRedGreen } from '../../src/events/pair';\n\ndescribe('pairRedGreen', () => {\n  it('should return the most recent fail→pass transition per test_id, ignoring test_ids that never passed', () => {\n    const events = [\n      { type: 'test', test_id: 't1', status: 'fail', ts: 1 },\n      { type: 'test', test_id: 't1', status: 'fail', ts: 2 },\n      { type: 'test', test_id: 't1', status: 'pass', ts: 3 },\n      { type: 'test', test_id: 't1', status: 'fail', ts: 4 },\n      { type: 'test', test_id: 't1', status: 'pass', ts: 5 },\n      { type: 'test', test_id: 't2', status: 'fail', ts: 10 },\n      { type: 'test', test_id: 't2', status: 'fail', ts: 11 },\n      { type: 'test', test_id: 't3', status: 'pass', ts: 20 },\n      { type: 'test', test_id: 't3', status: 'fail', ts: 21 },\n      { type: 'test', test_id: 't3', status: 'pass', ts: 22 },\n    ];\n\n    const result = pairRedGreen(events);\n\n    expect(result).toEqual([\n      { test_id: 't1', red_ts: 4, green_ts: 5 },\n      { test_id: 't3', red_ts: 21, green_ts: 22 },\n    ]);\n  });\n});"
-      }
-    }
-  ]
-}

benchmarks/spike-tdd-kernel/tdd-eval.md

@@ -1,81 +0,0 @@
-# Exp 3 — DeepSeek V4 TDD reliability
-
-**Result: PASS.** 8/10 strict, 10/10 once an over-strict scoring bug is corrected. Both thresholds (≥70% strict, ≥50% before redesign) cleared comfortably.
-
-## Method
-
-`benchmarks/spike-tdd-kernel/tdd-eval.mjs` runs 10 prompts across 5 easy / 3 medium / 2 hard difficulty levels against `deepseek-v4-flash` at temperature 0. The system message demands a single failing vitest file with no implementation. Each response is scored on:
-
-- **shape**: contains `describe`, `it`/`test`, and at least one `import`
-- **importsTarget**: imports the module-under-test by some path
-- **implLeak**: whether the test file defines the function-under-test (regression — the model was supposed to write only the test)
-- **stableNames**: every `it()`/`test()` title is a literal string with no template / `Date.now()` / `Math.random()`
-- **tsOk**: passes `tsc --noEmit` after replacing the target import with `vitest` (purely a syntax check)
-
-Pass-all requires all five.
-
-Raw runs in `tdd-eval.json` (~5 KB).
-
-## Numbers
-
-```
-e1 (easy)   shape=Y import=N* leak=N names=Y ts=Y → fail*
-e2 (easy)   shape=Y import=Y  leak=N names=Y ts=Y → PASS
-e3 (easy)   shape=Y import=Y  leak=N names=Y ts=Y → PASS
-e4 (easy)   shape=Y import=Y  leak=N names=Y ts=Y → PASS
-e5 (easy)   shape=Y import=Y  leak=N names=Y ts=Y → PASS
-m1 (medium) shape=Y import=Y  leak=N names=Y ts=Y → PASS
-m2 (medium) shape=Y import=Y  leak=N names=Y ts=Y → PASS
-m3 (medium) shape=Y import=Y  leak=N names=Y ts=Y → PASS
-h1 (hard)   shape=Y import=N* leak=N names=Y ts=Y → fail*
-h2 (hard)   shape=Y import=Y  leak=N names=Y ts=Y → PASS
-
-8/10 = 80% strict
-10/10 = 100% once import-path scoring is corrected (see below)
-tokens: 2246 prompt + 5732 completion (≈ $0.001 total)
-```
-
-## The two "failures" are scoring bugs
-
-The strict regex required imports of the form `from ".../src/util/slugify"`. The two failing prompts produced these imports:
-
-```
-e1: import { slugify } from '../slugify';                      // assumes test is co-located
-h1: import { extractTestId } from './src/repair/test-id';      // assumes test is project-root-relative
-```
-
-Both **import the correct symbol from a path that points at the right module**. They differ only in *where the test file is assumed to live*, which is a question the prompt didn't answer. In the real flow, the model also picks the test file location, so the import is self-consistent. These should count as PASS.
-
-## What the model got right consistently
-
-- 10/10 imported `vitest` correctly.
-- 10/10 wrote one `describe` + multiple `it` blocks, no nested test stubs.
-- 10/10 had stable, literal `it()` names — no parametrise leaks, no clocks, no RNG.
-- 10/10 did NOT define the target function in the test file (no impl leak).
-- 10/10 passed `tsc --noEmit` syntax check.
-- Median latency 8.2s, p95 14s. Slower than expected for `v4-flash`, but acceptable given output size (~500 tokens / response).
-
-## What the model got "wrong"
-
-- Underspec: when given no test file location, it guesses one. Reasonable behavior, but the kernel will need to specify (or accept the model's choice and write the file there).
-- Hard prompts (h1, h2) took 11–14s vs. easy ~5s. Acceptable.
-
-## Implications for the RFC
-
-1. **Greenfield flow is viable.** The model can reliably author a failing test first when explicitly told to. Open question §1 in RFC #25 can be closed: a structured `author_failing_test` tool is **not** required — a clear system message suffices.
-
-2. **The kernel should specify (or extract) the target test file path.** When `submit_plan` includes a step with `test_id`, it should also include `test_file_path`. The dispatcher uses that to:
-   - know where to write the failing test
-   - resolve the relative import path the model emits
-   - compute the eventual `<rel-path>::<fullName>` id
-
-3. **Strip-and-validate the model output.** Even though shape passed 10/10, the kernel should still:
-   - strip markdown fences (the model occasionally wraps in `\`\`\`ts ... \`\`\`` — none of the 10 did, but be defensive)
-   - reject any file that defines the target symbol (impl leak) before running it
-   - require the test fail with `Error: Cannot find module …` or a real assertion failure (not a SyntaxError)
-
-4. **Latency.** ~8s median per failing-test authoring. For a per-feature TDD step, that's fine. Combined with Exp 4's ~2s vitest run, the red event lands ≤12s after the user kicks off a feature — acceptable UX.
-
-## Decision
-
-Greenlight Exp 3. Combined with Exp 2 + Exp 4, the proposal's three feasibility risks are resolved. **Move to Exp 1.**

benchmarks/spike-tdd-kernel/tdd-eval.mjs

@@ -1,190 +0,0 @@
-// Exp 3 — does DeepSeek V4 reliably write a failing test FIRST?
-// Loads .env, runs N prompts asking for a vitest-style failing test only.
-// Scores each response on 4 axes and writes tdd-eval.json + tdd-eval.md.
-
-import { readFileSync, writeFileSync, mkdtempSync } from "node:fs";
-import { spawnSync } from "node:child_process";
-import { tmpdir } from "node:os";
-import { join } from "node:path";
-
-// Load .env manually (no dotenv dep in this repo).
-for (const line of readFileSync(new URL("../../.env", import.meta.url), "utf8").split(/\r?\n/)) {
-  const m = line.match(/^([A-Z_]+)\s*=\s*(.*)$/);
-  if (m && !process.env[m[1]]) process.env[m[1]] = m[2].replace(/^["']|["']$/g, "");
-}
-
-// Build DeepSeek client by importing the compiled dist (avoids tsx dep).
-// If dist is stale, fall back to direct fetch — same wire format.
-const apiKey = process.env.DEEPSEEK_API_KEY;
-const baseUrl = process.env.DEEPSEEK_BASE_URL ?? "https://api.deepseek.com/v1";
-if (!apiKey) {
-  console.error("DEEPSEEK_API_KEY missing from .env");
-  process.exit(1);
-}
-
-const SYSTEM = `You are writing a SINGLE failing vitest test. Strict rules:
-
-1. Output ONLY a TypeScript test file — no prose, no markdown fences, no implementation.
-2. The test MUST fail right now (the module / function it tests does not exist yet, or has the wrong behavior).
-3. Use exactly one top-level \`describe\` and one or more \`it()\` blocks. Do NOT include any function definitions other than the test bodies.
-4. Import the module-under-test using its expected final import path. The import will fail to resolve — that is correct, that is the red.
-5. Do NOT define stubs or fakes of the function-under-test inline. The test must reference the real (unimported / unimplemented) symbol.
-6. End with no trailing markdown.`;
-
-const PROMPTS = [
-  // easy (5)
-  { id: "e1", level: "easy", task: "A pure function \`slugify(s: string): string\` in src/util/slugify.ts that lowercases, replaces non-alphanumerics with '-', and collapses repeated dashes." },
-  { id: "e2", level: "easy", task: "A pure function \`clamp(n: number, lo: number, hi: number): number\` in src/util/clamp.ts that clamps n into [lo, hi]." },
-  { id: "e3", level: "easy", task: "A pure function \`hexToRgb(hex: string): {r:number,g:number,b:number} | null\` in src/util/color.ts. Accepts '#abc', '#aabbcc', and 'aabbcc'. Returns null on invalid." },
-  { id: "e4", level: "easy", task: "A pure function \`uniqueBy<T,K>(arr: T[], key: (t: T) => K): T[]\` in src/util/uniq.ts preserving first occurrence." },
-  { id: "e5", level: "easy", task: "A pure function \`parseDuration(s: string): number\` in src/util/duration.ts. '1500ms' → 1500, '2s' → 2000, '1m' → 60000. Returns NaN on invalid." },
-
-  // medium (3)
-  { id: "m1", level: "medium", task: "A class \`RingBuffer<T>\` in src/util/ring.ts with capacity, push(x) (drops oldest when full), toArray() returning oldest-first, and size getter." },
-  { id: "m2", level: "medium", task: "A function \`mergeRanges(ranges: Array<[number,number]>): Array<[number,number]>\` in src/util/ranges.ts. Coalesces overlapping/adjacent ranges, returns sorted." },
-  { id: "m3", level: "medium", task: "A function \`debounceAsync<T extends any[], R>(fn: (...args: T) => Promise<R>, ms: number): (...args: T) => Promise<R>\` in src/util/debounce.ts. Resolves only the latest call's promise; earlier callers reject with an AbortError-like." },
-
-  // hard (2) — these touch domain types from the repo
-  { id: "h1", level: "hard", task: "A function \`extractTestId(file: string, fullName: string, source: string): { id: string, source: 'native' | 'annotation' }\` in src/repair/test-id.ts. If \`source\` contains a '// @reasonix-test-id: <slug>' comment within 3 lines above an it()/test() whose name matches \`fullName\`, return that slug with source='annotation'. Otherwise return \`${file}::${fullName}\` with source='native'." },
-  { id: "h2", level: "hard", task: "A function \`pairRedGreen(events: Array<{type:string, test_id?:string, status?:string, ts:number}>): Array<{ test_id: string, red_ts: number, green_ts: number }>\` in src/events/pair.ts. For each test_id, find the most recent fail→pass transition and return one entry per test_id. Ignore test_ids that never went green." },
-];
-
-async function callModel(prompt) {
-  const body = {
-    model: "deepseek-v4-flash",
-    messages: [
-      { role: "system", content: SYSTEM },
-      { role: "user", content: prompt.task },
-    ],
-    temperature: 0.0,
-    max_tokens: 1500,
-    stream: false,
-  };
-  const t0 = Date.now();
-  const resp = await fetch(`${baseUrl}/chat/completions`, {
-    method: "POST",
-    headers: {
-      Authorization: `Bearer ${apiKey}`,
-      "Content-Type": "application/json",
-    },
-    body: JSON.stringify(body),
-  });
-  if (!resp.ok) {
-    const err = await resp.text();
-    throw new Error(`API ${resp.status}: ${err.slice(0, 300)}`);
-  }
-  const data = await resp.json();
-  const ms = Date.now() - t0;
-  const content = data.choices?.[0]?.message?.content ?? "";
-  return { content, ms, usage: data.usage };
-}
-
-function stripFences(s) {
-  return s
-    .replace(/^```(?:ts|typescript|tsx|javascript|js)?\s*\n/m, "")
-    .replace(/\n```\s*$/m, "")
-    .trim();
-}
-
-function score(prompt, raw) {
-  const code = stripFences(raw);
-
-  // (a) structurally a test file
-  const hasDescribe = /\bdescribe\s*\(/.test(code);
-  const hasIt = /\b(?:it|test)\s*\(/.test(code);
-  const hasImport = /^\s*import\s/m.test(code);
-  const compiles_shape = hasDescribe && hasIt && hasImport;
-
-  // (b) does it actually import the target module-under-test?
-  const targetMatch = prompt.task.match(/in (src\/[^\s.]+\.ts)/);
-  const target = targetMatch ? targetMatch[1].replace(/\.ts$/, "") : null;
-  const importsTarget = target
-    ? new RegExp(
-        `from\\s+["'](?:\\.\\.\\/(?:\\.\\.\\/)?)?(?:src\\/)?${target.replace("src/", "").replace(/\//g, "\\/")}`,
-      ).test(code)
-    : false;
-
-  // (c) impl leak — does the file define a function/class with the target's name?
-  const symbolMatch = prompt.task.match(/(?:function |class )?\\?`(\w+)/);
-  const symbol = symbolMatch ? symbolMatch[1] : null;
-  const implLeak =
-    symbol &&
-    new RegExp(
-      `(?:^|\\n)(?:export\\s+)?(?:function|class|const)\\s+${symbol}\\b\\s*[(=<{]`,
-    ).test(code);
-
-  // (d) at least one stable it() name (no template literals, no Date.now(), no RNG)
-  const itNames = [...code.matchAll(/\b(?:it|test)\s*\(\s*["'`]([^"'`]+)["'`]/g)].map(
-    (m) => m[1],
-  );
-  const stableNames = itNames.length > 0 && itNames.every((n) => !/\$\{|\bDate\.|Math\./.test(n));
-
-  // run typescript syntax-check via tsc on a temp file
-  let tsOk = false;
-  let tsErr = "";
-  try {
-    const dir = mkdtempSync(join(tmpdir(), "tdd-eval-"));
-    const f = join(dir, "candidate.test.ts");
-    // Replace the import path so tsc doesn't try to resolve it (we just want syntax + types of literals)
-    const stubbed = code.replace(/from\s+["'][^"']+["']/g, 'from "vitest"');
-    writeFileSync(f, stubbed);
-    const r = spawnSync(
-      "npx",
-      ["tsc", "--noEmit", "--target", "es2022", "--module", "esnext", "--moduleResolution", "bundler", "--skipLibCheck", "--strict", "false", f],
-      { encoding: "utf8", shell: true },
-    );
-    tsOk = r.status === 0 || /Cannot find module 'vitest'/i.test(r.stdout + r.stderr); // tolerate vitest miss
-    if (!tsOk) tsErr = (r.stdout + r.stderr).slice(0, 300);
-  } catch (e) {
-    tsErr = String(e).slice(0, 300);
-  }
-
-  const passAll =
-    compiles_shape && (importsTarget || target == null) && !implLeak && stableNames && tsOk;
-
-  return {
-    compiles_shape,
-    importsTarget,
-    implLeak: !!implLeak,
-    stableNames,
-    tsOk,
-    tsErr,
-    passAll,
-    target,
-    symbol,
-    itNames,
-    code,
-  };
-}
-
-console.log(`Running ${PROMPTS.length} prompts on deepseek-v4-flash …\n`);
-const out = [];
-let totalUsage = { prompt_tokens: 0, completion_tokens: 0 };
-for (const p of PROMPTS) {
-  process.stdout.write(`  ${p.id} (${p.level}) … `);
-  try {
-    const { content, ms, usage } = await callModel(p);
-    if (usage) {
-      totalUsage.prompt_tokens += usage.prompt_tokens ?? 0;
-      totalUsage.completion_tokens += usage.completion_tokens ?? 0;
-    }
-    const s = score(p, content);
-    out.push({ ...p, ms, usage, score: s });
-    console.log(
-      `${ms}ms  shape=${s.compiles_shape ? "Y" : "N"} import=${s.importsTarget ? "Y" : "N"} leak=${s.implLeak ? "Y" : "N"} names=${s.stableNames ? "Y" : "N"} ts=${s.tsOk ? "Y" : "N"} → ${s.passAll ? "PASS" : "fail"}`,
-    );
-  } catch (e) {
-    console.log(`ERROR: ${e.message}`);
-    out.push({ ...p, error: e.message });
-  }
-}
-
-const passed = out.filter((r) => r.score?.passAll).length;
-const total = out.length;
-console.log(`\n=== ${passed}/${total} pass-all (${((passed / total) * 100).toFixed(0)}%) ===`);
-console.log(`tokens: ${totalUsage.prompt_tokens} prompt + ${totalUsage.completion_tokens} completion`);
-
-writeFileSync(
-  new URL("./tdd-eval.json", import.meta.url),
-  JSON.stringify({ passed, total, totalUsage, runs: out }, null, 2),
-);

benchmarks/spike-tdd-kernel/test-id-spec.md

@@ -1,94 +0,0 @@
-# Exp 2 — `test_id` stability spec
-
-**Result: PASS (with hybrid).** Adopt vitest-native id (`<rel-path>::<fullName>`) as the default, with an optional annotation override for users who care about rename stability.
-
-## Schemes evaluated
-
-### A. vitest-native — `<relative-path>::<fullName>`
-
-Example: `tests/bang.test.ts::detectBangCommand returns the command body for a \`!\`-prefixed input`
-
-Verified against the JSON reporter (`npx vitest --reporter=json`); `fullName` is the documented `describe` chain joined with the leaf title and is what `-t "<fullName>"` matches against.
-
-| event | stable? |
-|---|---|
-| edit test **body** (logic, asserts) | yes |
-| rename `it()` title | **no** — id changes |
-| rename outer `describe()` | **no** — id changes |
-| move file | **no** — path changes |
-| reorder `describe` blocks | yes |
-| `it.each` parametrise: add row | yes for existing rows; new id appears |
-| `it.each` parametrise: change a row's args | id changes for that case |
-
-Critical failures: 3 (rename it / rename describe / move file).
-
-### B. content hash — sha256 of test body
-
-| event | stable? |
-|---|---|
-| edit test body | **no** — id changes on any whitespace edit |
-| rename it/describe | yes |
-| move file | yes |
-| parametrise | yes (body unchanged) |
-
-Critical failure: 1, but it's the worst possible one. Tests evolve while red — adding asserts, narrowing scope. A scheme that invalidates `test_id` on every body edit makes `edit_claim` impossible to track across the red→green journey. **Reject.**
-
-### C. user annotation — `// @reasonix-test-id: foo`
-
-| event | stable? |
-|---|---|
-| edit body / rename / move | yes |
-| parametrise | ambiguous — one id, N runs |
-| greenfield | requires model to invent + uniqueness-check |
-| existing 96 test files | zero have it; brownfield bootstrap is awkward |
-
-Critical failures: 2 (parametrise ambiguity, brownfield bootstrap). Strong on rename, weak on adoption.
-
-## Decision: hybrid (A as default, C as opt-in override)
-
-Default `test_id` = `<rel-path>::<fullName>`.
-If the test source contains `// @reasonix-test-id: <slug>` directly above the `it(`/`test(`, that slug overrides the default.
-
-```ts
-// @reasonix-test-id: bang.parses-leading-bang
-it('returns the command body for a `!`-prefixed input', () => { … });
-```
-
-This handles the failure modes of A:
-- **Rename it/describe**: a user who anticipates renames adds the annotation once. Without it, kernel treats rename as a new test (correct — the old red is gone, so should be the old claim).
-- **Move file**: same — annotation makes the id survive moves.
-- **Brownfield**: zero churn for existing 96 files; they use the default.
-- **Greenfield**: model uses the default unless the user requests stability. `reasonix doctor` could surface a warning when a `test_id` would be lost.
-
-### How the dispatcher resolves it
-
-When extracting `test_id` from a `test_run` event, the kernel:
-1. Parses `--reporter=json` output → `{file, fullName}`.
-2. Reads the test source (already in workspace).
-3. If an annotation comment within 3 lines above the matched `it(` exists, use the slug.
-4. Else use `<rel-path>::<fullName>`.
-
-This is deterministic and replayable from `events.jsonl` alone (the source at the time of the event is captured by the workspace snapshot).
-
-## Implications for the RFC
-
-Update RFC §"New event types":
-
-```ts
-type TestRunEvent = {
-  type: 'test_run';
-  test_id: string;           // <rel-path>::<fullName>  OR  user annotation slug
-  test_id_source: 'native' | 'annotation';   // for debugging / migration
-  status: 'pass' | 'fail';
-  command: string;
-  duration_ms: number;
-  ts: number;
-};
-```
-
-Add §"`test_id` resolution" subsection citing this spec.
-
-## Out of scope (defer)
-
-- Cross-runner support (jest, mocha). Reasonix workspaces today are predominantly vitest; ship vitest-only first.
-- Refactor-safe id (e.g., AST-based fingerprint resilient to whitespace + rename). Possible v2.

benchmarks/spike-tdd-kernel/work-estimate.md

@@ -1,86 +0,0 @@
-# Staged work estimate — kernel red-green (RFC #25)
-
-> Local-only estimate, paired with `tracking-issue-draft.md`. Numbers are wall-clock for one focused day, not "ideal" hours.
-
-## Total
-
-~4–5 days of actual coding across all three stages. Then ~2 minor releases of soak before flipping default-on.
-
-| Stage | Code (LoC) | Tests (LoC) | Wall time | Risk |
-|---|---|---|---|---|
-| 1. events + writer | ~300 | ~150 | 0.5 day | low |
-| 2. dispatcher gate | ~600 | ~400 | 2–3 days | **high** |
-| 3. plan + UI | ~250 | ~120 | 1 day | medium |
-
-## Stage 1 — events + writer (0.5 day)
-
-Almost entirely additive, no behavior change.
-
-**Changes:**
-- `src/core/events.ts:190` — extend `Event` union with `TestRunEvent` + `EditClaimEvent`.
-- `src/core/test-id.ts` (new, ~50 LoC) — `extractTestId(file, fullName, source)` per `test-id-spec.md`.
-- `src/core/reducers/red-green.ts` (new, ~30 LoC) — `pairRedGreen(events)` reducer.
-- `src/cli/commands/events.ts` — add `red-green` subcommand listing pairs.
-- `src/adapters/event-sink-jsonl.ts` — already generic over `Event`, no edits required.
-
-**Tests:**
-- Round-trip: append a `test_run` event, replay through reducer.
-- `extractTestId` matrix: 8 cases (rename, move, parametrise, annotation override, etc.).
-
-**Risk:** low. Pattern matches existing event additions in v0.14.
-
-## Stage 2 — dispatcher gate (2–3 days, **the load-bearing one**)
-
-This is where most of the actual integration risk lives.
-
-**Changes:**
-- `src/tools/filesystem.ts:518` — `edit_file` registration wraps in a gate. When `REASONIX_STRICT_TDD=1`:
-  1. Look up most recent `test_run` for `test_id` from in-memory event list (cheaper than re-reading jsonl).
-  2. Verify a matching `edit_claim` followed it.
-  3. On dispatch refusal, throw a structured error the model can read.
-- `src/loop.ts` — per-turn coalescing buffer:
-  - When `edit_file` succeeds, push `{test_id, test_file_path}` to a turn-scoped Set.
-  - At end-of-turn (just before the next assistant call), spawn one `vitest --run -t a -t b -t c` covering all collected ids.
-  - Parse `--reporter=json` output, emit one `test_run` event per id.
-  - On any red, revert the offending edits via the existing checkpoint mechanism (`src/checkpoints.ts`), emit a `repair` event so the storm-breaker engages.
-- `/refactor` mode — session flag in `LoopState`. When true, gate is bypassed; on session exit, run `npm run verify` (or `reasonix.config.ts`'s `verify_command`).
-- `reasonix.config.ts` schema — add `verify_command` and `test_command_for(test_id)`.
-
-**Tests:**
-- Integration on a synthetic session fixture: green path, red revert, multi-edit batch, `/refactor` bypass, edit before any test_id (refused).
-- Mock vitest spawner so tests don't depend on actual vitest runs.
-
-**Risk: high.** Specific concerns:
-- **Loop coordination.** End-of-turn flush has to play nice with: abort controller (`_turnAbort`), /pro escalation (mid-turn model swap), storm-breaker (`src/repair/storm.ts`), thinking-mode round-trip (reasoning_content preservation). Any one of these can desynchronise the buffer.
-- **Vitest spawn hang.** Need timeout + kill + emit a `test_run` event with `status='fail'` and a tagged failure reason. Otherwise a stuck test hangs the whole agent.
-- **Cross-platform paths.** Vitest's `fullName` should be POSIX-normalised before becoming part of `test_id`; spike runs were on Windows but didn't stress this.
-- **Revert semantics.** If batch had 3 edits and 1 went red, only that file reverts; others stay. Existing `Checkpoint` is per-file, but the index (`src/checkpoints.ts`) needs a partial-restore code path.
-
-**Mitigation:** land stage 2 in two PRs — first the gate + buffer behind a new flag (no auto-run), then the auto-run + revert. Validates the synchronisation before adding the spawner.
-
-## Stage 3 — plan + UI (1 day)
-
-**Changes:**
-- `src/tools/plan-types.ts:3` — `PlanStep` gains `test_id?` + `test_file_path?`.
-- `src/tools/plan-core.ts` — `submit_plan` validation: any step with `test_id` must have `test_file_path`.
-- `src/cli/commands/doctor.ts` — warn when plan has `test_id` but missing `test_file_path`; warn on first session in an untested codebase, suggest `/refactor` default.
-- TUI plan card — render red/green dots per step (need to inspect `src/cli/ui/cards/PlanCard*` to see how steps render today).
-
-**Tests:**
-- Plan validation: rejects step with `test_id` missing `test_file_path`.
-- Doctor output: snapshot of warning lines.
-- TUI snapshot for a 3-step plan with mixed red/green/pending dots.
-
-**Risk: medium.** TUI rendering is the unknown — depends on whether the current plan card has slots for status badges, or if the layout needs widening.
-
-## Default-on rollout (calendar, not work)
-
-- After stage 3 lands: minor release with flag *off* by default.
-- Two minor releases of soak — collect any hangs / false-refusals via telemetry, fix in patches.
-- Flip default-on; keep `REASONIX_STRICT_TDD=0` opt-out for two more minor releases.
-
-## Cross-cutting risks not pinned to a stage
-
-1. **Untested codebases.** `reasonix doctor` should detect (no `tests/` dir, no `vitest.config.*`) and refuse to enable strict mode at all on first run. Otherwise the flag is unusable.
-2. **Greenfield test-file location.** Spike Exp 3 showed the model picks reasonable but inconsistent paths when none is specified. The plan-step `test_file_path` field is the fix, but a user editing a single file with no plan still has the gap. Stage 2 should refuse `edit_file` when strict + no `test_file_path` is in scope.
-3. **MCP-served edit tools.** Reasonix supports MCP-hosted tools (`src/mcp.ts`). If an MCP server exposes its own write/edit tool, the kernel gate doesn't apply. Stage 2 should at minimum log a warning; longer-term, MCP write tools could opt into the same gate via a hook.

benchmarks/tau-bench/baseline.ts

@@ -1,144 +0,0 @@
-/** Naive baseline — deliberately breaks prefix cache (fresh timestamp + shuffled tool keys + full-rebuild log) so the comparison vs CacheFirstLoop isolates Pillar 1. */
-
-import {
-  type ChatMessage,
-  type DeepSeekClient,
-  SessionStats,
-  type ToolCall,
-  type ToolDefinition,
-  ToolRegistry,
-  type ToolSpec,
-  type Usage,
-} from "../../src/index.js";
-import type { Turn } from "./types.js";
-
-export interface BaselineRunnerOptions {
-  client: DeepSeekClient;
-  systemPrompt: string;
-  tools: ToolDefinition[];
-  model?: string;
-  maxToolIters?: number;
-}
-
-export interface BaselineSubCall {
-  /** Assistant text from this sub-call (often empty when the response is tool-calls-only). */
-  content: string;
-  /** Usage for this single client.chat() call. */
-  usage: Usage;
-  /** Tools the model chose to call on the back of this response. */
-  toolCalls: { name: string; args: string; result: string }[];
-}
-
-export interface BaselineTurnResult {
-  assistantMessage: string;
-  toolCallsExecuted: { name: string; args: string; result: string }[];
-  /** Per-sub-call breakdown so bench transcripts match Reasonix loop-event granularity. */
-  subCalls: BaselineSubCall[];
-  /** Turn number (1-based) assigned by the agent. */
-  turnNo: number;
-}
-
-export class BaselineAgent {
-  readonly client: DeepSeekClient;
-  readonly stats = new SessionStats();
-  private readonly systemPrompt: string;
-  private readonly registry: ToolRegistry;
-  private readonly model: string;
-  private readonly maxToolIters: number;
-  /** Previous-turn messages — kept, but the prefix rebuilds around them every turn so cache churns. */
-  private history: ChatMessage[] = [];
-  private turnNo = 0;
-
-  constructor(opts: BaselineRunnerOptions) {
-    this.client = opts.client;
-    this.systemPrompt = opts.systemPrompt;
-    this.model = opts.model ?? "deepseek-chat";
-    this.maxToolIters = opts.maxToolIters ?? 6;
-    this.registry = new ToolRegistry({ autoFlatten: false });
-    for (const t of opts.tools) this.registry.register(t);
-  }
-
-  /** Run one user-turn — intentionally non-cache-friendly (fresh ts + shuffled tool specs every turn). */
-  async userTurn(userMessage: string, transcript: Turn[]): Promise<BaselineTurnResult> {
-    this.turnNo++;
-
-    // Naive pattern #1: current-time placeholder in the system prompt.
-    const churnedSystem = `${this.systemPrompt}\nCurrent time: ${new Date().toISOString()}`;
-
-    // Naive pattern #2: shuffle tool spec order each turn (simulates
-    // frameworks that materialize tools from Python dicts / maps).
-    const shuffledTools = shuffle(this.registry.specs(), this.turnNo);
-
-    this.history.push({ role: "user", content: userMessage });
-
-    const toolExecutions: { name: string; args: string; result: string }[] = [];
-    const subCalls: BaselineSubCall[] = [];
-
-    for (let iter = 0; iter < this.maxToolIters; iter++) {
-      // Naive pattern #3: always rebuild the full message array.
-      const messages: ChatMessage[] = [{ role: "system", content: churnedSystem }, ...this.history];
-
-      const resp = await this.client.chat({
-        model: this.model,
-        messages,
-        tools: shuffledTools,
-      });
-      this.stats.record(this.turnNo, this.model, resp.usage);
-
-      const assistantMessage: ChatMessage = { role: "assistant", content: resp.content };
-      if (resp.toolCalls.length > 0) assistantMessage.tool_calls = resp.toolCalls;
-      this.history.push(assistantMessage);
-
-      if (resp.toolCalls.length === 0) {
-        subCalls.push({ content: resp.content, usage: resp.usage, toolCalls: [] });
-        return {
-          assistantMessage: resp.content,
-          toolCallsExecuted: toolExecutions,
-          subCalls,
-          turnNo: this.turnNo,
-        };
-      }
-
-      const subToolCalls: { name: string; args: string; result: string }[] = [];
-      for (const tc of resp.toolCalls) {
-        const name = tc.function?.name ?? "";
-        const args = tc.function?.arguments ?? "{}";
-        const result = await this.registry.dispatch(name, args);
-        toolExecutions.push({ name, args, result });
-        subToolCalls.push({ name, args, result });
-        this.history.push({
-          role: "tool",
-          tool_call_id: tc.id ?? "",
-          name,
-          content: result,
-        });
-      }
-      subCalls.push({ content: resp.content, usage: resp.usage, toolCalls: subToolCalls });
-    }
-
-    const lastAssistant = [...this.history].reverse().find((m) => m.role === "assistant");
-    return {
-      assistantMessage: lastAssistant?.content ?? "[max_tool_iters reached]",
-      toolCallsExecuted: toolExecutions,
-      subCalls,
-      turnNo: this.turnNo,
-    };
-  }
-}
-
-/**
- * Deterministic Fisher–Yates seeded by turn-number — reproducible runs, cache-hostile orderings.
- */
-function shuffle<T>(arr: T[], seed: number): T[] {
-  const out = [...arr];
-  let s = seed * 9301 + 49297;
-  for (let i = out.length - 1; i > 0; i--) {
-    s = (s * 9301 + 49297) % 233280;
-    const j = Math.floor((s / 233280) * (i + 1));
-    [out[i], out[j]] = [out[j]!, out[i]!];
-  }
-  return out;
-}
-
-// Re-export ToolCall, ToolSpec so caller files don't need to import both places.
-export type { ToolCall, ToolSpec };

benchmarks/tau-bench/db.ts

@@ -1,30 +0,0 @@
-/** tau-bench in-memory DB helpers — pure functions over a nested-table WorldState. */
-import type { WorldState } from "./types.js";
-
-/** Deep-clone a WorldState — `structuredClone` is enough since the type is JSON-shaped by contract. Idempotent: calling cloneDb N times on the same input yields identical copies. */
-export function cloneDb(db: WorldState): WorldState {
-  return structuredClone(db);
-}
-
-/** Get a row by table name and id. Lookup is `db[table][id]` — two-level nested dictionary. Returns the row object or `undefined` if not found. */
-export function getRow(
-  db: WorldState,
-  table: string,
-  id: string,
-): Record<string, unknown> | undefined {
-  return db[table]?.[id];
-}
-
-/** Set a single field on a row. Mutates `db` in-place — caller should `cloneDb` first if isolation is needed. Returns `true` if the row was found and updated, `false` if not. */
-export function setField(
-  db: WorldState,
-  table: string,
-  id: string,
-  field: string,
-  value: unknown,
-): boolean {
-  const row = db[table]?.[id];
-  if (!row) return false;
-  row[field] = value;
-  return true;
-}

benchmarks/tau-bench/report.md

@@ -1,100 +0,0 @@
-# Reasonix tool-use eval (τ-bench-lite)
-
-**Date:** 2026-04-29T16:34:53.893Z
-**Agent model:** `deepseek-chat`
-**User-simulator model:** `deepseek-chat`
-**Tasks:** 8, repeats × 3
-**Reasonix version:** 0.16.0
-
-## Summary
-
-| metric | baseline | reasonix | delta |
-|---|---:|---:|---:|
-| runs | 24 | 24 | — |
-| pass rate | 100% | 100% | +0pp |
-| cache hit | 32.8% | 90.2% | **+57.4pp** |
-| mean cost / task | $0.000992 | $0.000593 | ×0.60 |
-| mean turns | 4.8 | 4.3 | — |
-| mean tool calls | 2.7 | 2.7 | — |
-
-**Reasonix vs Claude Sonnet 4.6 (estimated, same token counts):**
-Claude would cost ~$0.039998 / task, so Reasonix saves ~98.1%.
-(This is a *token-count-based estimate*, not a head-to-head quality comparison.)
-
-## Per-task breakdown
-
-| task | mode | pass | turns | tools | cache | cost |
-|---|---|:---:|---:|---:|---:|---:|
-| t01_address_happy | baseline | ✅ | 3 | 3 | 47.9% | $0.000579 |
-| t01_address_happy | reasonix | ✅ | 2 | 3 | 88.6% | $0.000329 |
-| t01_address_happy | baseline | ✅ | 3 | 3 | 46.4% | $0.000577 |
-| t01_address_happy | reasonix | ✅ | 3 | 3 | 91.0% | $0.000383 |
-| t01_address_happy | baseline | ✅ | 3 | 3 | 38.7% | $0.000538 |
-| t01_address_happy | reasonix | ✅ | 3 | 3 | 91.4% | $0.000381 |
-| t02_address_not_allowed | baseline | ✅ | 8 | 2 | 6.6% | $0.001809 |
-| t02_address_not_allowed | reasonix | ✅ | 8 | 3 | 91.9% | $0.001170 |
-| t02_address_not_allowed | baseline | ✅ | 8 | 2 | 7.0% | $0.001644 |
-| t02_address_not_allowed | reasonix | ✅ | 8 | 2 | 90.0% | $0.001021 |
-| t02_address_not_allowed | baseline | ✅ | 8 | 2 | 12.5% | $0.001788 |
-| t02_address_not_allowed | reasonix | ✅ | 7 | 2 | 90.6% | $0.000891 |
-| t03_cancel_processing | baseline | ✅ | 2 | 3 | 59.4% | $0.000412 |
-| t03_cancel_processing | reasonix | ✅ | 2 | 3 | 86.3% | $0.000321 |
-| t03_cancel_processing | baseline | ✅ | 2 | 3 | 59.6% | $0.000409 |
-| t03_cancel_processing | reasonix | ✅ | 3 | 3 | 90.6% | $0.000360 |
-| t03_cancel_processing | baseline | ✅ | 2 | 3 | 60.0% | $0.000408 |
-| t03_cancel_processing | reasonix | ✅ | 2 | 3 | 93.2% | $0.000291 |
-| t04_refund_delivered | baseline | ✅ | 3 | 3 | 49.2% | $0.000598 |
-| t04_refund_delivered | reasonix | ✅ | 3 | 3 | 93.5% | $0.000379 |
-| t04_refund_delivered | baseline | ✅ | 3 | 3 | 47.6% | $0.000599 |
-| t04_refund_delivered | reasonix | ✅ | 2 | 3 | 91.1% | $0.000320 |
-| t04_refund_delivered | baseline | ✅ | 3 | 3 | 48.7% | $0.000608 |
-| t04_refund_delivered | reasonix | ✅ | 2 | 3 | 93.5% | $0.000335 |
-| t05_refund_not_delivered | baseline | ✅ | 8 | 2 | 7.1% | $0.001631 |
-| t05_refund_not_delivered | reasonix | ✅ | 7 | 2 | 89.0% | $0.000990 |
-| t05_refund_not_delivered | baseline | ✅ | 8 | 2 | 7.0% | $0.001686 |
-| t05_refund_not_delivered | reasonix | ✅ | 8 | 3 | 93.3% | $0.001294 |
-| t05_refund_not_delivered | baseline | ✅ | 6 | 3 | 22.3% | $0.001295 |
-| t05_refund_not_delivered | reasonix | ✅ | 7 | 2 | 89.7% | $0.000878 |
-| t06_multi_order_lookup | baseline | ✅ | 4 | 2 | 26.8% | $0.000726 |
-| t06_multi_order_lookup | reasonix | ✅ | 4 | 2 | 87.5% | $0.000478 |
-| t06_multi_order_lookup | baseline | ✅ | 4 | 2 | 25.5% | $0.000798 |
-| t06_multi_order_lookup | reasonix | ✅ | 3 | 2 | 84.9% | $0.000332 |
-| t06_multi_order_lookup | baseline | ✅ | 4 | 2 | 28.1% | $0.000748 |
-| t06_multi_order_lookup | reasonix | ✅ | 3 | 2 | 88.0% | $0.000398 |
-| t07_wrong_identity | baseline | ✅ | 8 | 2 | 12.0% | $0.001686 |
-| t07_wrong_identity | reasonix | ✅ | 8 | 2 | 88.7% | $0.001066 |
-| t07_wrong_identity | baseline | ✅ | 8 | 2 | 11.7% | $0.001734 |
-| t07_wrong_identity | reasonix | ✅ | 4 | 2 | 88.1% | $0.000573 |
-| t07_wrong_identity | baseline | ✅ | 8 | 2 | 12.5% | $0.001629 |
-| t07_wrong_identity | reasonix | ✅ | 6 | 2 | 87.7% | $0.000843 |
-| t08_address_then_cancel | baseline | ✅ | 3 | 4 | 48.6% | $0.000629 |
-| t08_address_then_cancel | reasonix | ✅ | 2 | 3 | 94.1% | $0.000307 |
-| t08_address_then_cancel | baseline | ✅ | 3 | 4 | 49.4% | $0.000603 |
-| t08_address_then_cancel | reasonix | ✅ | 3 | 4 | 89.7% | $0.000468 |
-| t08_address_then_cancel | baseline | ✅ | 3 | 4 | 52.7% | $0.000677 |
-| t08_address_then_cancel | reasonix | ✅ | 3 | 4 | 92.8% | $0.000434 |
-
-## Scope & caveats
-
-This is **τ-bench-lite**, not a port of Sierra's upstream τ-bench. Specifically:
-
-- Tasks are hand-authored in the retail domain; the schema mirrors τ-bench
-  (stateful tools, LLM user-sim, DB-end-state success predicates), so upstream
-  tasks can later be dropped in without harness changes.
-- Every pass/fail judgment is a deterministic DB predicate — no LLM judge.
-  Refusal tasks pass iff the DB is unchanged.
-- The "baseline" deliberately reproduces cache-hostile patterns common in
-  generic agent frameworks: fresh timestamp in the system prompt each turn,
-  re-shuffled tool spec ordering per turn. It is **not** a benchmark of
-  LangChain specifically.
-- Claude comparison is a *token-count-based cost estimate* using Anthropic's
-  public pricing, not a head-to-head quality run.
-- User simulator is DeepSeek V3 at T=0.1. Some run-to-run drift is expected;
-  rerun with `--repeats N` to get a tighter mean.
-
-## Reproducing
-
-1. `export DEEPSEEK_API_KEY=sk-...`
-2. `npm install`
-3. `npx tsx benchmarks/tau-bench/runner.ts --repeats 3`
-4. `npx tsx benchmarks/tau-bench/report.ts benchmarks/tau-bench/results-*.json`

benchmarks/tau-bench/report.ts

@@ -1,218 +0,0 @@
-/** Render τ-bench results.json → report.md. CLI usage in benchmarks/README.md. */
-
-import { readFileSync, writeFileSync } from "node:fs";
-import { pathToFileURL } from "node:url";
-import type { BenchReport, RunMode, RunResult } from "./types.js";
-
-interface CliArgs {
-  input: string;
-  outPath: string;
-}
-
-function parseArgs(argv: string[]): CliArgs {
-  const out: CliArgs = { input: "", outPath: "benchmarks/tau-bench/report.md" };
-  const positional: string[] = [];
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i];
-    if (a === "--out") out.outPath = argv[++i] ?? out.outPath;
-    else if (a && !a.startsWith("--")) positional.push(a);
-  }
-  out.input = positional[0] ?? "";
-  if (!out.input) {
-    throw new Error(
-      "usage: npx tsx benchmarks/tau-bench/report.ts <results.json> [--out report.md]",
-    );
-  }
-  return out;
-}
-
-interface Agg {
-  runs: number;
-  passes: number;
-  avgCache: number;
-  avgCost: number;
-  avgClaudeCost: number;
-  avgTurns: number;
-  avgToolCalls: number;
-}
-
-function aggregate(results: RunResult[]): Agg {
-  if (results.length === 0) {
-    return {
-      runs: 0,
-      passes: 0,
-      avgCache: 0,
-      avgCost: 0,
-      avgClaudeCost: 0,
-      avgTurns: 0,
-      avgToolCalls: 0,
-    };
-  }
-  const passes = results.filter((r) => r.pass).length;
-  const mean = (fn: (r: RunResult) => number) =>
-    results.reduce((s, r) => s + fn(r), 0) / results.length;
-  return {
-    runs: results.length,
-    passes,
-    avgCache: mean((r) => r.cacheHitRatio),
-    avgCost: mean((r) => r.costUsd),
-    avgClaudeCost: mean((r) => r.claudeEquivalentUsd),
-    avgTurns: mean((r) => r.turns),
-    avgToolCalls: mean((r) => r.toolCalls),
-  };
-}
-
-function renderSummary(report: BenchReport): string {
-  const byMode: Record<RunMode, RunResult[]> = { baseline: [], reasonix: [] };
-  for (const r of report.results) byMode[r.mode].push(r);
-  const b = aggregate(byMode.baseline);
-  const rx = aggregate(byMode.reasonix);
-
-  const costRatio = b.avgCost > 0 ? rx.avgCost / b.avgCost : 0;
-  const claudeSavings = b.avgClaudeCost > 0 ? (1 - rx.avgCost / b.avgClaudeCost) * 100 : 0;
-
-  return `
-## Summary
-
-| metric | baseline | reasonix | delta |
-|---|---:|---:|---:|
-| runs | ${b.runs} | ${rx.runs} | — |
-| pass rate | ${pct(b.passes, b.runs)} | ${pct(rx.passes, rx.runs)} | ${signPct(rx.passes, rx.runs, b.passes, b.runs)} |
-| cache hit | ${pct1(b.avgCache)} | ${pct1(rx.avgCache)} | **${signPctAbs(rx.avgCache - b.avgCache)}** |
-| mean cost / task | $${fmt(b.avgCost, 6)} | $${fmt(rx.avgCost, 6)} | ${costRatio > 0 ? `×${fmt(costRatio, 2)}` : "—"} |
-| mean turns | ${fmt(b.avgTurns, 1)} | ${fmt(rx.avgTurns, 1)} | — |
-| mean tool calls | ${fmt(b.avgToolCalls, 1)} | ${fmt(rx.avgToolCalls, 1)} | — |
-
-**Reasonix vs Claude Sonnet 4.6 (estimated, same token counts):**
-Claude would cost ~$${fmt(rx.avgClaudeCost, 6)} / task, so Reasonix saves ~${fmt(
-    claudeSavings,
-    1,
-  )}%.
-(This is a *token-count-based estimate*, not a head-to-head quality comparison.)
-`.trim();
-}
-
-function renderPerTask(report: BenchReport): string {
-  const rowsByTask = new Map<string, RunResult[]>();
-  for (const r of report.results) {
-    const list = rowsByTask.get(r.taskId) ?? [];
-    list.push(r);
-    rowsByTask.set(r.taskId, list);
-  }
-  const rows: string[] = [
-    "| task | mode | pass | turns | tools | cache | cost |",
-    "|---|---|:---:|---:|---:|---:|---:|",
-  ];
-  for (const [taskId, runs] of rowsByTask) {
-    for (const r of runs) {
-      rows.push(
-        `| ${taskId} | ${r.mode} | ${r.pass ? "✅" : "❌"} | ${r.turns} | ${r.toolCalls} | ${pct1(
-          r.cacheHitRatio,
-        )} | $${fmt(r.costUsd, 6)} |${r.errorMessage ? ` err: ${truncate(r.errorMessage, 40)}` : ""}`,
-      );
-    }
-  }
-  return `## Per-task breakdown\n\n${rows.join("\n")}`;
-}
-
-function renderHeader(report: BenchReport): string {
-  const m = report.meta;
-  return `# Reasonix tool-use eval (τ-bench-lite)
-
-**Date:** ${m.date}
-**Agent model:** \`${m.model}\`
-**User-simulator model:** \`${m.userSimModel}\`
-**Tasks:** ${m.taskCount}, repeats × ${m.repeatsPerTask}
-**Reasonix version:** ${m.reasonixVersion}
-`;
-}
-
-function renderCaveats(): string {
-  return `## Scope & caveats
-
-This is **τ-bench-lite**, not a port of Sierra's upstream τ-bench. Specifically:
-
-- Tasks are hand-authored in the retail domain; the schema mirrors τ-bench
-  (stateful tools, LLM user-sim, DB-end-state success predicates), so upstream
-  tasks can later be dropped in without harness changes.
-- Every pass/fail judgment is a deterministic DB predicate — no LLM judge.
-  Refusal tasks pass iff the DB is unchanged.
-- The "baseline" deliberately reproduces cache-hostile patterns common in
-  generic agent frameworks: fresh timestamp in the system prompt each turn,
-  re-shuffled tool spec ordering per turn. It is **not** a benchmark of
-  LangChain specifically.
-- Claude comparison is a *token-count-based cost estimate* using Anthropic's
-  public pricing, not a head-to-head quality run.
-- User simulator is DeepSeek V3 at T=0.1. Some run-to-run drift is expected;
-  rerun with \`--repeats N\` to get a tighter mean.
-
-## Reproducing
-
-1. \`export DEEPSEEK_API_KEY=sk-...\`
-2. \`npm install\`
-3. \`npx tsx benchmarks/tau-bench/runner.ts --repeats 3\`
-4. \`npx tsx benchmarks/tau-bench/report.ts benchmarks/tau-bench/results-*.json\`
-`;
-}
-
-export function renderReport(report: BenchReport): string {
-  return [
-    renderHeader(report),
-    renderSummary(report),
-    "",
-    renderPerTask(report),
-    "",
-    renderCaveats(),
-  ].join("\n");
-}
-
-function pct(num: number, denom: number): string {
-  if (denom === 0) return "—";
-  return `${((num / denom) * 100).toFixed(0)}%`;
-}
-
-function signPct(num1: number, denom1: number, num2: number, denom2: number): string {
-  if (denom1 === 0 || denom2 === 0) return "—";
-  const d = num1 / denom1 - num2 / denom2;
-  const s = (d * 100).toFixed(0);
-  return `${d >= 0 ? "+" : ""}${s}pp`;
-}
-
-function signPctAbs(diff: number): string {
-  const s = (diff * 100).toFixed(1);
-  return `${diff >= 0 ? "+" : ""}${s}pp`;
-}
-
-function pct1(x: number): string {
-  return `${(x * 100).toFixed(1)}%`;
-}
-
-function fmt(x: number, digits: number): string {
-  return x.toFixed(digits);
-}
-
-function truncate(s: string, n: number): string {
-  return s.length > n ? `${s.slice(0, n)}…` : s;
-}
-
-async function main(): Promise<void> {
-  const args = parseArgs(process.argv.slice(2));
-  const raw = readFileSync(args.input, "utf8");
-  const report = JSON.parse(raw) as BenchReport;
-  const md = renderReport(report);
-  writeFileSync(args.outPath, md, "utf8");
-  console.log(`wrote ${args.outPath} (${report.results.length} runs)`);
-}
-
-function isMain(): boolean {
-  const entry = process.argv[1];
-  if (!entry) return false;
-  return import.meta.url === pathToFileURL(entry).href;
-}
-
-if (isMain()) {
-  main().catch((err) => {
-    console.error(err);
-    process.exit(1);
-  });
-}

benchmarks/tau-bench/results.json

@@ -1,684 +0,0 @@
-{
-  "meta": {
-    "date": "2026-04-29T16:34:53.893Z",
-    "model": "deepseek-chat",
-    "userSimModel": "deepseek-chat",
-    "taskCount": 8,
-    "repeatsPerTask": 3,
-    "reasonixVersion": "0.16.0"
-  },
-  "results": [
-    {
-      "taskId": "t01_address_happy",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.4792298092351578,
-      "costUsd": 0.000579139,
-      "claudeEquivalentUsd": 0.022106999999999998,
-      "promptTokens": 5609,
-      "completionTokens": 352,
-      "truncated": false,
-      "finalAgentMessage": "Done! The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help with?"
-    },
-    {
-      "taskId": "t01_address_happy",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.8861538461538462,
-      "costUsd": 0.000328628,
-      "claudeEquivalentUsd": 0.021929999999999998,
-      "promptTokens": 5200,
-      "completionTokens": 422,
-      "truncated": false,
-      "finalAgentMessage": "Done! Your shipping address for order **o_1002** has been updated to:\n\n**5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t01_address_happy",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.46410442349528647,
-      "costUsd": 0.000576806,
-      "claudeEquivalentUsd": 0.021633,
-      "promptTokens": 5516,
-      "completionTokens": 339,
-      "truncated": false,
-      "finalAgentMessage": "Done! The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t01_address_happy",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9103277674706246,
-      "costUsd": 0.0003825380000000001,
-      "claudeEquivalentUsd": 0.026799000000000003,
-      "promptTokens": 6468,
-      "completionTokens": 493,
-      "truncated": false,
-      "finalAgentMessage": "The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t01_address_happy",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.38654012079378774,
-      "costUsd": 0.0005375100000000002,
-      "claudeEquivalentUsd": 0.018873,
-      "promptTokens": 4636,
-      "completionTokens": 331,
-      "truncated": false,
-      "finalAgentMessage": "All done! The shipping address for order **o_1002** has been updated to:\n\n> **5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t01_address_happy",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9142857142857143,
-      "costUsd": 0.00038142600000000003,
-      "claudeEquivalentUsd": 0.026864999999999997,
-      "promptTokens": 6440,
-      "completionTokens": 503,
-      "truncated": false,
-      "finalAgentMessage": "Your shipping address has been updated successfully! Here's a summary:\n\n- **Order ID:** o_1002\n- **Item:** running shoes 10\n- **Old Address:** 22 Oak Rd, NYC, NY 10001\n- **New Address:** 5 Birch Rd, NYC, NY 10001\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.06629261976694001,
-      "costUsd": 0.001808905,
-      "claudeEquivalentUsd": 0.05007,
-      "promptTokens": 11585,
-      "completionTokens": 1021,
-      "truncated": true,
-      "finalAgentMessage": "I understand your frustration, and I'm sorry \u2014 but I don't have any tools to add notes to your account or record requests in the system. So unfortunately, I can't confirm that this request has been logged anywhere.\n\nIf you reach out to the main support team (by email or phone), I'd recommend asking them to add a note to your account about the address change request. That way it's on record.\n\nAgain, really sorry I couldn't help more with this, Ari. Let me know if there's anything else I can assist with!"
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9194998903268261,
-      "costUsd": 0.0011700640000000002,
-      "claudeEquivalentUsd": 0.08149800000000001,
-      "promptTokens": 18236,
-      "completionTokens": 1786,
-      "truncated": true,
-      "finalAgentMessage": "I understand you'd like a supervisor, but I must be straightforward with you \u2014 **I have no mechanism to transfer, escalate, or connect you to anyone else**. Repeating the same request will not change what I'm capable of doing.\n\nYour options remain:\n\n1. **Contact customer support** (phone/email) to request a supervisor.\n2. **Reroute with the shipping carrier** once you have tracking info.\n3. **Wait for delivery** and then we can discuss a refund or return.\n\nI want to help, but this is the extent of what I can do. Is there anything else I can assist you with?"
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.07032967032967033,
-      "costUsd": 0.001644468,
-      "claudeEquivalentUsd": 0.04419,
-      "promptTokens": 10920,
-      "completionTokens": 762,
-      "truncated": true,
-      "finalAgentMessage": "I appreciate you asking, but I'm not able to confirm that \u2014 I don't have any tools to log, record, or store notes on your account or order. So unfortunately, nothing has been recorded in the system on my end.\n\nYour best bet is to contact the shipping carrier directly or reach out via the website's contact page. Thank you for your patience, Ari. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.9000067608680955,
-      "costUsd": 0.001021171,
-      "claudeEquivalentUsd": 0.068268,
-      "promptTokens": 14791,
-      "completionTokens": 1593,
-      "truncated": true,
-      "finalAgentMessage": "I understand your disappointment, and I'm sorry. Unfortunately, I can't confirm that anything has been logged \u2014 I don't have any ticketing, note-taking, or recording tools on my end. Your request hasn't been stored anywhere through me.\n\nHere's a quick summary of what I'd recommend you take with you:\n\n\ud83d\udcdd **Contact main support with:**\n- **Order:** o_1001\n- **Current address:** 1 Elm St, SF, CA 94110\n- **New address:** 99 New St, SF, CA\n- **Request:** Reroute with carrier (since it's already shipped)\n\nAgain, I wish I could do more. I hope you get it resolved quickly, Ari."
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.12467532467532468,
-      "costUsd": 0.001788292,
-      "claudeEquivalentUsd": 0.05025,
-      "promptTokens": 12320,
-      "completionTokens": 886,
-      "truncated": true,
-      "finalAgentMessage": "I appreciate your persistence, but I've told you everything I'm capable of. I'm not going to be able to change the address or transfer you, no matter how many times you ask \u2014 my system simply won't allow it.\n\nPlease check our website for customer service contact details. I hope you get this sorted out, Ari! Take care."
-    },
-    {
-      "taskId": "t02_address_not_allowed",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 7,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.9055675176868656,
-      "costUsd": 0.0008907320000000001,
-      "claudeEquivalentUsd": 0.06007199999999999,
-      "promptTokens": 13004,
-      "completionTokens": 1404,
-      "truncated": false,
-      "finalAgentMessage": "I completely understand your frustration, Ari, and I'm sorry. To be totally transparent with you \u2014 I only have the tools you've seen here, and I don't have a directory, transfer system, or any way to connect you to a specific person or department. I can't say for sure if someone else could do it, but I don't have a way to get you to them directly.\n\nThe general **[email protected]** or the phone number on your order confirmation is likely your best path forward. I hope they can resolve this for you \u2014 best of luck!"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.5939019001325674,
-      "costUsd": 0.00041220000000000004,
-      "claudeEquivalentUsd": 0.017973000000000003,
-      "promptTokens": 4526,
-      "completionTokens": 293,
-      "truncated": false,
-      "finalAgentMessage": "All done, Dev! Your order **o_1004** (kettle) has been successfully cancelled. No trouble at all \u2014 if you need anything else, feel free to ask!"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.8629213483146068,
-      "costUsd": 0.000321073,
-      "claudeEquivalentUsd": 0.020595000000000002,
-      "promptTokens": 4895,
-      "completionTokens": 394,
-      "truncated": false,
-      "finalAgentMessage": "Your order **o_1004** has been successfully cancelled. You're all set! Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.5960088691796009,
-      "costUsd": 0.00040886400000000005,
-      "claudeEquivalentUsd": 0.017865,
-      "promptTokens": 4510,
-      "completionTokens": 289,
-      "truncated": false,
-      "finalAgentMessage": "All done! Your order **o_1004** (kettle) has been successfully **canceled**. You're welcome, Dev! Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9060489060489061,
-      "costUsd": 0.00035980200000000005,
-      "claudeEquivalentUsd": 0.025173,
-      "promptTokens": 6216,
-      "completionTokens": 435,
-      "truncated": false,
-      "finalAgentMessage": "Your order **o_1004** has been successfully cancelled! \u2705 \n\nYou're all set, Dev. If there's anything else I can help you with, feel free to ask. \ud83d\ude0a"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.5995984831585992,
-      "costUsd": 0.00040816900000000006,
-      "claudeEquivalentUsd": 0.017949,
-      "promptTokens": 4483,
-      "completionTokens": 300,
-      "truncated": false,
-      "finalAgentMessage": "Your order **o_1004** (kettle) has been successfully cancelled. No worries at all, happy to help! Is there anything else I can assist you with?"
-    },
-    {
-      "taskId": "t03_cancel_processing",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9316619490497372,
-      "costUsd": 0.00029082,
-      "claudeEquivalentUsd": 0.021032999999999996,
-      "promptTokens": 4946,
-      "completionTokens": 413,
-      "truncated": false,
-      "finalAgentMessage": "Your order **o_1004** has been successfully canceled! \ud83c\udf89 Since it was still in processing, no charges should have been finalized \u2014 but if you have any concerns about a refund or anything else, feel free to ask.\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.49213561691716184,
-      "costUsd": 0.0005984300000000001,
-      "claudeEquivalentUsd": 0.023406,
-      "promptTokens": 5722,
-      "completionTokens": 416,
-      "truncated": false,
-      "finalAgentMessage": "The refund has been processed successfully! Here's a summary:\n\n- **Order**: o_1003 (Desk Lamp)\n- **Refund Amount**: $55 (full refund)\n- **Reason**: Item arrived damaged (shattered glass shade)\n\nYou should see the refund reflected in your payment method soon. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9346031746031747,
-      "costUsd": 0.00037864600000000006,
-      "claudeEquivalentUsd": 0.027345,
-      "promptTokens": 6300,
-      "completionTokens": 563,
-      "truncated": false,
-      "finalAgentMessage": "The refund has been issued successfully! Here's a summary:\n\n- **Order:** o_1003 (Desk Lamp)\n- **Refund Amount:** $55.00\n- **Reason:** Item arrived damaged\n- **Status:** \u2705 Refund processed\n\nThe amount should be credited back to your original payment method within a few business days. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.47592067988668557,
-      "costUsd": 0.000599294,
-      "claudeEquivalentUsd": 0.023019,
-      "promptTokens": 5648,
-      "completionTokens": 405,
-      "truncated": false,
-      "finalAgentMessage": "The refund has been successfully processed! Here's a summary:\n\n- **Order:** o_1003 (Desk Lamp)\n- **Refund Amount:** $55 \u2014 **Full refund** \u2705\n- **Reason:** Item arrived damaged\n\nYou should see the refund reflected in your original payment method within a few business days. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9109394062627084,
-      "costUsd": 0.00032004,
-      "claudeEquivalentUsd": 0.021969000000000002,
-      "promptTokens": 4918,
-      "completionTokens": 481,
-      "truncated": false,
-      "finalAgentMessage": "I've successfully issued a **full refund of $55** for your desk lamp (order o_1003) due to the damage. The refund has been processed. You should see it reflected in your original payment method soon.\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.4865238424326192,
-      "costUsd": 0.00060816,
-      "claudeEquivalentUsd": 0.023634,
-      "promptTokens": 5788,
-      "completionTokens": 418,
-      "truncated": false,
-      "finalAgentMessage": "All done! The **$55 refund** for your desk lamp (order **o_1003**) has been issued successfully due to the damage. The refund should process back to your original payment method.\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t04_refund_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9348598499802606,
-      "costUsd": 0.00033471400000000006,
-      "claudeEquivalentUsd": 0.023628,
-      "promptTokens": 5066,
-      "completionTokens": 562,
-      "truncated": false,
-      "finalAgentMessage": "All done! \u2705 A **full refund of $55** has been issued for your desk lamp (order **o_1003**) due to the damage. You should see the amount reflected in your original payment method within a few business days.\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.07105847520355292,
-      "costUsd": 0.0016314020000000001,
-      "claudeEquivalentUsd": 0.04398900000000001,
-      "promptTokens": 10808,
-      "completionTokens": 771,
-      "truncated": true,
-      "finalAgentMessage": "I understand you feel strongly about this, but I've been transparent about what I can and can't do. I cannot process a refund on o_1002 right now due to its current status. \n\nYou're welcome to take this higher \u2014 that might be the best path forward if you want an option beyond what I can offer. In the meantime, if you change your mind about a **cancellation**, feel free to reach back out."
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 7,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8897036358081271,
-      "costUsd": 0.000990008,
-      "claudeEquivalentUsd": 0.064266,
-      "promptTokens": 13092,
-      "completionTokens": 1666,
-      "truncated": false,
-      "finalAgentMessage": "Bo, I understand you're insisting, but I cannot do something that's outside the capability of my tools. I've told you everything available to me, and nothing has changed. I'm not holding back \u2014 there simply isn't an option to refund a processing order.\n\nI'm happy to help with anything I *can* do (cancel, update address, answer questions), but I can't keep responding to the same request when the answer won't change. Is there anything else I can assist you with?"
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.0702588967157625,
-      "costUsd": 0.0016863070000000004,
-      "claudeEquivalentUsd": 0.046398,
-      "promptTokens": 10931,
-      "completionTokens": 907,
-      "truncated": true,
-      "finalAgentMessage": "I've been honest with you about what I can do, and I can't magically create capabilities I don't have. Continuing to ask the same thing won't change that.\n\nTo summarize your real options one last time:\n\n- **Cancel** order o_1002 \u2192 money back, done now.\n- **Do nothing** \u2192 the order ships and you can request a refund once delivered.\n\nIf you'd like to cancel, just let me know. Otherwise, I'd suggest contacting the company via their phone or email support to speak with a supervisor. I'm unable to do anything further from here."
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9325418240690772,
-      "costUsd": 0.0012944879999999999,
-      "claudeEquivalentUsd": 0.08995499999999998,
-      "promptTokens": 18530,
-      "completionTokens": 2291,
-      "truncated": true,
-      "finalAgentMessage": "I hear you, Bo, and I completely get why you're upset. But even with the threat of a bad review, I can't make the system do something it doesn't allow \u2014 that would be dishonest. I've already tried the refund and it was rejected.\n\nHere's a quick recap so you don't have to repeat yourself when you reach someone who can escalate:\n\n- **Order:** o_1002 \u2013 Running shoes 10 \u2013 $140\n- **Status:** Processing\n- **Issue:** You want a refund now; system requires delivered status\n- **Expedient option:** Cancel now for an immediate refund\n- **Other option:** Wait for delivery, then get your refund\n\nI'm sorry I couldn't fully resolve this for you today. If you decide to go with cancellation or need anything else, I'm here."
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 6,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.22306938242021565,
-      "costUsd": 0.0012951390000000003,
-      "claudeEquivalentUsd": 0.040833,
-      "promptTokens": 9181,
-      "completionTokens": 886,
-      "truncated": false,
-      "finalAgentMessage": "I completely understand you're upset, and I want to be straightforward with you. I'm **not refusing to help** \u2014 I actually attempted the refund and the system rejected it with an error because the order is still in processing status. I've exhausted every tool I have.\n\nI don't have an escalation or supervisor transfer tool available to me. I'm limited to what's in front of me.\n\nTo summarize:\n- **Refund**: \u274c System rejected it (order not yet delivered)\n- **Cancel**: \u2705 I can do this right now if you want\n- **Address change**: \u2705 I can update it if needed\n- **Escalation**: \u274c Not possible from here\n\nIf you'd like to cancel the order to stop it from shipping and prevent charges from going through, just say the word and I'll do it immediately. Otherwise, I'd suggest contacting support via another channel to request a supervisor."
-    },
-    {
-      "taskId": "t05_refund_not_delivered",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 7,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8973153827770254,
-      "costUsd": 0.000877647,
-      "claudeEquivalentUsd": 0.058329,
-      "promptTokens": 12553,
-      "completionTokens": 1378,
-      "truncated": false,
-      "finalAgentMessage": "I hear you, and I understand you're upset. But I have to be honest \u2014 I don't have access to any internal directories, phone numbers, or email contacts for supervisors or departments. I can't make up a number or pretend to transfer you. That would be dishonest.\n\nI've given you all the options my system allows. If you'd like to pursue a chargeback, that's your right. \n\nIs there anything else within my ability I can help you with?"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 4,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.26843760922754284,
-      "costUsd": 0.000725776,
-      "claudeEquivalentUsd": 0.022611,
-      "promptTokens": 5722,
-      "completionTokens": 363,
-      "truncated": false,
-      "finalAgentMessage": "That's right! Both orders have moved past the \"processing\" stage:\n\n- **o_1001** \u2014 Shipped\n- **o_1005** \u2014 Delivered\n\nSo neither can be canceled anymore. No problem at all \u2014 happy to help anytime! Is there anything else I can do for you?"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 4,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8753799392097265,
-      "costUsd": 0.00047764400000000005,
-      "claudeEquivalentUsd": 0.03066,
-      "promptTokens": 6580,
-      "completionTokens": 728,
-      "truncated": false,
-      "finalAgentMessage": "Sure! However, looking at your orders:\n\n- **o_1001** (Wool Sweater) \u2014 **Shipped** (already on its way, can't cancel)\n- **o_1005** (Notebook Pack) \u2014 **Delivered** (already delivered, can't cancel)\n\nNeither is in \"processing\" status, so there's nothing to cancel. Your sweater is already on its way! Anything else I can help with?"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 4,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.25523429710867396,
-      "costUsd": 0.000798334,
-      "claudeEquivalentUsd": 0.025193999999999998,
-      "promptTokens": 6018,
-      "completionTokens": 476,
-      "truncated": false,
-      "finalAgentMessage": "Absolutely! I've already seen the full list of your orders \u2014 there are only **two**:\n\n1. **o_1001** \u2013 **Shipped** (not processing)\n2. **o_1005** \u2013 **Delivered** (not processing)\n\nNeither is in \"processing\" status, so no cancellations can be made. You're all set! Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8493487343327599,
-      "costUsd": 0.00033237300000000007,
-      "claudeEquivalentUsd": 0.020322,
-      "promptTokens": 4069,
-      "completionTokens": 541,
-      "truncated": false,
-      "finalAgentMessage": "Those are all the orders linked to your account (**u_ari**) \u2014 just the two I listed. No other orders are in processing or anywhere else on file.\n\nSo nothing's missing! The wool sweater is shipped and on its way, and the notebook pack has already been delivered. Let me know if you need anything else! \ud83d\ude0a"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 4,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.2808438818565401,
-      "costUsd": 0.0007481250000000001,
-      "claudeEquivalentUsd": 0.02367,
-      "promptTokens": 5925,
-      "completionTokens": 393,
-      "truncated": false,
-      "finalAgentMessage": "You're welcome, Ari! Glad everything looks good. Have a great day! \ud83d\ude0a"
-    },
-    {
-      "taskId": "t06_multi_order_lookup",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8803974016048911,
-      "costUsd": 0.000398406,
-      "claudeEquivalentUsd": 0.025542,
-      "promptTokens": 5234,
-      "completionTokens": 656,
-      "truncated": false,
-      "finalAgentMessage": "I've already pulled up your full order history, and those are **all the orders** linked to your account (u_ari). Unfortunately, there aren't any orders with a \"processing\" status \u2014 just the shipped and delivered ones we saw.\n\nIt's possible that a more recent order might have been placed under a different account or email. Would you like to check with another user ID or order number? Otherwise, if there's anything I can help with regarding the existing orders, just say the word!"
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.12033159559012051,
-      "costUsd": 0.001686157,
-      "claudeEquivalentUsd": 0.046758,
-      "promptTokens": 11701,
-      "completionTokens": 777,
-      "truncated": true,
-      "finalAgentMessage": "I understand you're persistent, but I've checked the system thoroughly:\n\n- **Order o_1004** \u2192 user: **Dev Patel**\n- **Your name**: **Bo Wang**\n\nThese do not match. I will not cancel the order. I'm happy to help if you provide a correct user ID or order ID that matches your name. Otherwise, I've exhausted my options here."
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8873635802045439,
-      "costUsd": 0.001066297,
-      "claudeEquivalentUsd": 0.06940199999999999,
-      "promptTokens": 14569,
-      "completionTokens": 1713,
-      "truncated": true,
-      "finalAgentMessage": "I've answered this multiple times now. To be clear one last time:\n\n- **I cannot cancel order o_1004** because the system shows it belongs to **Dev Patel**, not Bo Wang.\n- I will not process this request regardless of how many times it's repeated.\n\nIf you'd like to try a different approach \u2014 such as verifying through the registered email or providing a different order number linked to your name \u2014 I'm here to help. Otherwise, I'm closing this thread."
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.11737245748582861,
-      "costUsd": 0.001733556,
-      "claudeEquivalentUsd": 0.047987999999999996,
-      "promptTokens": 11996,
-      "completionTokens": 800,
-      "truncated": true,
-      "finalAgentMessage": "I'm sorry, but I've answered this several times now and my answer won't change. Let me help you move forward:\n\n1. Do you have a **user ID** I can look you up with?\n2. Or a **different order ID** that might be under your name, Bo Wang?\n\nIf neither of those works, please contact our customer support team who can help resolve any account discrepancies. Is there anything else I can assist you with?"
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 4,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8812392426850258,
-      "costUsd": 0.000573186,
-      "claudeEquivalentUsd": 0.036351,
-      "promptTokens": 6972,
-      "completionTokens": 1029,
-      "truncated": false,
-      "finalAgentMessage": "I appreciate your patience, but I really can't make changes without proper identity verification. To proceed with a cancellation, I need either:\n\n1. **Your user ID** (so I can confirm your name matches the order), or\n2. **An order ID that is actually registered under Bo Wang**\n\nRight now, the system shows o_1004 belongs to Dev Patel, not Bo Wang. If you believe there's been a mix-up, please contact our customer support team with proof of purchase so they can investigate further.\n\nIs there anything else I can help you with?"
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 8,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.12492236713690001,
-      "costUsd": 0.0016294450000000002,
-      "claudeEquivalentUsd": 0.04563300000000001,
-      "promptTokens": 11271,
-      "completionTokens": 788,
-      "truncated": true,
-      "finalAgentMessage": "I understand you're upset, but I simply cannot override the system. To be transparent, here's what I see on my end:\n\n| Detail | Value |\n|---|---|\n| Order ID | o_1004 |\n| Item | Kettle |\n| Address | 4 Maple Ln, Seattle, WA 98101 |\n| Registered Name | **Dev Patel** |\n| Email | [email protected] |\n\nThe name mismatch is a red flag I can't ignore. If **you are Dev Patel**, please let me know. Otherwise, please find your **user ID** (from any previous email or account page) and I'll help you right away."
-    },
-    {
-      "taskId": "t07_wrong_identity",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 6,
-      "toolCalls": 2,
-      "cacheHitRatio": 0.8768815129293709,
-      "costUsd": 0.0008429900000000001,
-      "claudeEquivalentUsd": 0.053277,
-      "promptTokens": 10364,
-      "completionTokens": 1479,
-      "truncated": false,
-      "finalAgentMessage": "I understand you feel strongly about this, but I've exhausted the options available to me. My tools cannot override the identity mismatch, and repeating the same request won't change that.\n\nPlease contact our **customer support team** through another channel (phone or email) so they can investigate the account discrepancy. That's the only way forward here."
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 4,
-      "cacheHitRatio": 0.4857284276522026,
-      "costUsd": 0.0006288410000000001,
-      "claudeEquivalentUsd": 0.024288,
-      "promptTokens": 6061,
-      "completionTokens": 407,
-      "truncated": false,
-      "finalAgentMessage": "Done! Order **o_1004** (the kettle) has been cancelled successfully. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 2,
-      "toolCalls": 3,
-      "cacheHitRatio": 0.9406444318824194,
-      "costUsd": 0.00030699300000000003,
-      "claudeEquivalentUsd": 0.022580999999999997,
-      "promptTokens": 5307,
-      "completionTokens": 444,
-      "truncated": false,
-      "finalAgentMessage": "Done! Order **o_1004** has been successfully cancelled. The shipping address update is no longer needed. Is there anything else I can help you with?"
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 4,
-      "cacheHitRatio": 0.49429147078576224,
-      "costUsd": 0.000603126,
-      "claudeEquivalentUsd": 0.023372999999999998,
-      "promptTokens": 5956,
-      "completionTokens": 367,
-      "truncated": false,
-      "finalAgentMessage": "Order **o_1004** has been cancelled successfully. Is there anything else I can help with?"
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 4,
-      "cacheHitRatio": 0.8970091352771868,
-      "costUsd": 0.000467723,
-      "claudeEquivalentUsd": 0.032208,
-      "promptTokens": 7991,
-      "completionTokens": 549,
-      "truncated": false,
-      "finalAgentMessage": "Order **o_1004** has been cancelled successfully. Let me know if you need anything else!"
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "baseline",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 4,
-      "cacheHitRatio": 0.5269733106189665,
-      "costUsd": 0.000676616,
-      "claudeEquivalentUsd": 0.027042000000000004,
-      "promptTokens": 7044,
-      "completionTokens": 394,
-      "truncated": false,
-      "finalAgentMessage": "Done! Order **o_1004** has been cancelled. Is there anything else I can help with?"
-    },
-    {
-      "taskId": "t08_address_then_cancel",
-      "mode": "reasonix",
-      "pass": true,
-      "turns": 3,
-      "toolCalls": 4,
-      "cacheHitRatio": 0.928,
-      "costUsd": 0.000434164,
-      "claudeEquivalentUsd": 0.031889999999999995,
-      "promptTokens": 8000,
-      "completionTokens": 526,
-      "truncated": false,
-      "finalAgentMessage": "Order **o_1004** has been cancelled successfully. Is there anything else I can help with?"
-    }
-  ]
-}

benchmarks/tau-bench/runner.ts

@@ -1,444 +0,0 @@
-/** τ-bench-lite runner — writes results.json. CLI flags + sample invocations in benchmarks/README.md. */
-
-import { type WriteStream, mkdirSync, writeFileSync } from "node:fs";
-import { join } from "node:path";
-import { pathToFileURL } from "node:url";
-import {
-  CacheFirstLoop,
-  DeepSeekClient,
-  ImmutablePrefix,
-  ToolRegistry,
-  VERSION,
-  claudeEquivalentCost,
-  costUsd,
-  loadDotenv,
-} from "../../src/index.js";
-import { openTranscriptFile, recordFromLoopEvent, writeRecord } from "../../src/transcript/log.js";
-import { BaselineAgent } from "./baseline.js";
-import { cloneDb } from "./db.js";
-import { TASKS } from "./tasks.js";
-import type { BenchReport, RunMode, RunResult, TaskDefinition, Turn, WorldState } from "./types.js";
-import { UserSimulator } from "./user-sim.js";
-
-loadDotenv();
-
-interface CliArgs {
-  taskFilter: string | null;
-  modes: RunMode[];
-  repeats: number;
-  model: string;
-  userSimModel: string;
-  outPath: string | null;
-  transcriptsDir: string | null;
-  dry: boolean;
-  verbose: boolean;
-}
-
-function parseArgs(argv: string[]): CliArgs {
-  const out: CliArgs = {
-    taskFilter: null,
-    modes: ["baseline", "reasonix"],
-    repeats: 1,
-    model: "deepseek-chat",
-    userSimModel: "deepseek-chat",
-    outPath: null,
-    transcriptsDir: null,
-    dry: false,
-    verbose: false,
-  };
-  for (let i = 0; i < argv.length; i++) {
-    const a = argv[i];
-    if (a === "--task") out.taskFilter = argv[++i] ?? null;
-    else if (a === "--mode") {
-      const v = (argv[++i] ?? "").toLowerCase();
-      if (v === "baseline" || v === "reasonix") out.modes = [v];
-    } else if (a === "--repeats") out.repeats = Number.parseInt(argv[++i] ?? "1", 10);
-    else if (a === "--model") out.model = argv[++i] ?? out.model;
-    else if (a === "--user-model") out.userSimModel = argv[++i] ?? out.userSimModel;
-    else if (a === "--out") out.outPath = argv[++i] ?? null;
-    else if (a === "--transcripts-dir") out.transcriptsDir = argv[++i] ?? null;
-    else if (a === "--dry") out.dry = true;
-    else if (a === "--verbose" || a === "-v") out.verbose = true;
-  }
-  return out;
-}
-
-interface RunContext {
-  client: DeepSeekClient;
-  task: TaskDefinition;
-  db: WorldState;
-  transcript: Turn[];
-  args: CliArgs;
-  /** Open transcript stream, or null if --transcripts-dir was not set. */
-  transcriptStream: WriteStream | null;
-}
-
-/** Convert a task's tool factories into concrete ToolDefinitions bound to this run's db. */
-function buildTools(task: TaskDefinition, db: WorldState) {
-  return task.tools.map((factory) => factory(db));
-}
-
-async function runReasonix(ctx: RunContext): Promise<RunResult> {
-  const { client, task, db, args, transcriptStream } = ctx;
-  const tools = buildTools(task, db);
-  const registry = new ToolRegistry();
-  for (const t of tools) registry.register(t);
-
-  const prefix = new ImmutablePrefix({
-    system: task.systemPrompt,
-    toolSpecs: registry.specs(),
-  });
-  const loop = new CacheFirstLoop({
-    client,
-    prefix,
-    tools: registry,
-    model: args.model,
-    stream: false,
-  });
-  const prefixHash = prefix.fingerprint;
-
-  return runAgentLoop(ctx, "reasonix", async (userMsg) => {
-    const stepTurns: Turn[] = [];
-    let finalText = "";
-    for await (const ev of loop.step(userMsg)) {
-      if (transcriptStream && ev.role !== "assistant_delta") {
-        writeRecord(transcriptStream, recordFromLoopEvent(ev, { model: args.model, prefixHash }));
-      }
-      if (ev.role === "tool") {
-        stepTurns.push({ role: "tool", content: ev.content, toolName: ev.toolName });
-      } else if (ev.role === "assistant_final") {
-        finalText = ev.content;
-      } else if (ev.role === "done") {
-        finalText = ev.content || finalText;
-        break;
-      } else if (ev.role === "error") {
-        throw new Error(ev.error ?? "reasonix loop error");
-      }
-    }
-    return {
-      assistantMessage: finalText,
-      toolEvents: stepTurns,
-      cacheHitRatio: loop.stats.aggregateCacheHitRatio,
-      costUsd: loop.stats.totalCost,
-      claudeEquivalentUsd: loop.stats.totalClaudeEquivalent,
-      promptTokens: sumTokens(loop.stats.turns.map((t) => t.usage.promptTokens)),
-      completionTokens: sumTokens(loop.stats.turns.map((t) => t.usage.completionTokens)),
-    };
-  });
-}
-
-async function runBaseline(ctx: RunContext): Promise<RunResult> {
-  const { client, task, db, args, transcriptStream } = ctx;
-  const tools = buildTools(task, db);
-  const agent = new BaselineAgent({
-    client,
-    systemPrompt: task.systemPrompt,
-    tools,
-    model: args.model,
-  });
-
-  return runAgentLoop(ctx, "baseline", async (userMsg, transcript) => {
-    const res = await agent.userTurn(userMsg, transcript);
-
-    // Emit one assistant_final + its tool records per sub-call, mirroring
-    // Reasonix's per-model-call granularity. This keeps diff apples-to-
-    // apples: a sub-call in baseline corresponds to one model call, which
-    // is also how Reasonix counts.
-    if (transcriptStream) {
-      for (const sc of res.subCalls) {
-        const ts = new Date().toISOString();
-        writeRecord(transcriptStream, {
-          ts,
-          turn: res.turnNo,
-          role: "assistant_final",
-          content: sc.content,
-          usage: {
-            prompt_tokens: sc.usage.promptTokens,
-            completion_tokens: sc.usage.completionTokens,
-            total_tokens: sc.usage.totalTokens,
-            prompt_cache_hit_tokens: sc.usage.promptCacheHitTokens,
-            prompt_cache_miss_tokens: sc.usage.promptCacheMissTokens,
-          },
-          cost: costUsd(args.model, sc.usage),
-          model: args.model,
-          // No prefixHash: baseline's prefix churns by design.
-        });
-        for (const tc of sc.toolCalls) {
-          writeRecord(transcriptStream, {
-            ts,
-            turn: res.turnNo,
-            role: "tool",
-            content: tc.result,
-            tool: tc.name,
-            args: tc.args,
-          });
-        }
-      }
-    }
-
-    return {
-      assistantMessage: res.assistantMessage,
-      toolEvents: res.toolCallsExecuted.map(
-        (x) => ({ role: "tool", content: x.result, toolName: x.name }) as Turn,
-      ),
-      cacheHitRatio: agent.stats.aggregateCacheHitRatio,
-      costUsd: agent.stats.totalCost,
-      claudeEquivalentUsd: agent.stats.turns.reduce((s, t) => s + claudeEquivalentCost(t.usage), 0),
-      promptTokens: sumTokens(agent.stats.turns.map((t) => t.usage.promptTokens)),
-      completionTokens: sumTokens(agent.stats.turns.map((t) => t.usage.completionTokens)),
-    };
-  });
-}
-
-interface AgentTurnOutput {
-  assistantMessage: string;
-  toolEvents: Turn[];
-  cacheHitRatio: number;
-  costUsd: number;
-  claudeEquivalentUsd: number;
-  promptTokens: number;
-  completionTokens: number;
-}
-
-async function runAgentLoop(
-  ctx: RunContext,
-  mode: RunMode,
-  userTurnFn: (userMsg: string, transcript: Turn[]) => Promise<AgentTurnOutput>,
-): Promise<RunResult> {
-  const { client, task, db, transcript, args, transcriptStream } = ctx;
-  const sim = new UserSimulator(client, task.user, {
-    model: args.userSimModel,
-    temperature: 0.1,
-  });
-
-  const maxTurns = task.maxTurns ?? 8;
-  let turns = 0;
-  let toolCalls = 0;
-  let lastAgentMsg = "";
-  let truncated = false;
-  let lastAgentOutput: AgentTurnOutput | null = null;
-  let errorMessage: string | undefined;
-
-  try {
-    while (turns < maxTurns) {
-      const userMsg = await sim.next(transcript);
-      if (userMsg === null) break;
-      transcript.push({ role: "user", content: userMsg });
-      if (transcriptStream) {
-        writeRecord(transcriptStream, {
-          ts: new Date().toISOString(),
-          turn: turns + 1,
-          role: "user",
-          content: userMsg,
-        });
-      }
-      if (args.verbose) console.log(`  [${mode}] USER: ${userMsg}`);
-
-      const out = await userTurnFn(userMsg, transcript);
-      lastAgentOutput = out;
-      for (const te of out.toolEvents) {
-        transcript.push(te);
-        toolCalls++;
-        if (args.verbose) console.log(`  [${mode}] TOOL ${te.toolName}: ${truncate(te.content)}`);
-      }
-      transcript.push({ role: "agent", content: out.assistantMessage });
-      if (args.verbose) console.log(`  [${mode}] AGENT: ${truncate(out.assistantMessage)}`);
-      lastAgentMsg = out.assistantMessage;
-      turns++;
-    }
-    if (turns === maxTurns) truncated = true;
-  } catch (err) {
-    errorMessage = (err as Error).message;
-  }
-
-  const pass =
-    errorMessage === undefined
-      ? safeCheck(task, { db, finalAgentMessage: lastAgentMsg, transcript })
-      : false;
-
-  return {
-    taskId: task.id,
-    mode,
-    pass,
-    turns,
-    toolCalls,
-    cacheHitRatio: lastAgentOutput?.cacheHitRatio ?? 0,
-    costUsd: lastAgentOutput?.costUsd ?? 0,
-    claudeEquivalentUsd: lastAgentOutput?.claudeEquivalentUsd ?? 0,
-    promptTokens: lastAgentOutput?.promptTokens ?? 0,
-    completionTokens: lastAgentOutput?.completionTokens ?? 0,
-    truncated,
-    finalAgentMessage: lastAgentMsg,
-    errorMessage,
-  };
-}
-
-function safeCheck(
-  task: TaskDefinition,
-  ctx: { db: WorldState; finalAgentMessage: string; transcript: Turn[] },
-): boolean {
-  try {
-    return task.check(ctx);
-  } catch {
-    return false;
-  }
-}
-
-function sumTokens(arr: number[]): number {
-  return arr.reduce((s, n) => s + n, 0);
-}
-
-function truncate(s: string, n = 140): string {
-  return s.length > n ? `${s.slice(0, n)}…` : s;
-}
-
-async function runDry(args: CliArgs): Promise<BenchReport> {
-  const tasks = filterTasks(args.taskFilter);
-  const results: RunResult[] = [];
-  for (const task of tasks) {
-    for (const mode of args.modes) {
-      const db = cloneDb(task.initialDb);
-      const tools = buildTools(task, db);
-      // Execute each tool once with dummy-ish args just to prove wiring works.
-      for (const t of tools) {
-        try {
-          await t.fn(stubArgs(t));
-        } catch {
-          /* dry mode — ignore tool errors */
-        }
-      }
-      results.push({
-        taskId: task.id,
-        mode,
-        pass: true,
-        turns: 0,
-        toolCalls: tools.length,
-        cacheHitRatio: mode === "reasonix" ? 0.9 : 0.1,
-        costUsd: 0,
-        claudeEquivalentUsd: 0,
-        promptTokens: 0,
-        completionTokens: 0,
-        truncated: false,
-        finalAgentMessage: "[dry-run]",
-      });
-      console.log(`[${task.id}/${mode}] dry-run ok (${tools.length} tools wired)`);
-    }
-  }
-  return {
-    meta: buildMeta(args, tasks.length),
-    results,
-  };
-}
-
-function stubArgs(t: {
-  name: string;
-  parameters?: { properties?: Record<string, unknown> };
-}): unknown {
-  const props = t.parameters?.properties ?? {};
-  const out: Record<string, unknown> = {};
-  for (const k of Object.keys(props)) out[k] = "stub";
-  return out;
-}
-
-function filterTasks(filter: string | null): TaskDefinition[] {
-  if (!filter) return TASKS;
-  const t = TASKS.find((x) => x.id === filter);
-  if (!t) throw new Error(`unknown task: ${filter}`);
-  return [t];
-}
-
-function buildMeta(args: CliArgs, taskCount: number): BenchReport["meta"] {
-  return {
-    date: new Date().toISOString(),
-    model: args.model,
-    userSimModel: args.userSimModel,
-    taskCount,
-    repeatsPerTask: args.repeats,
-    reasonixVersion: VERSION,
-  };
-}
-
-async function main(): Promise<void> {
-  const args = parseArgs(process.argv.slice(2));
-
-  if (args.dry) {
-    const report = await runDry(args);
-    writeReport(report, args.outPath);
-    return;
-  }
-
-  const client = new DeepSeekClient();
-  const tasks = filterTasks(args.taskFilter);
-
-  if (args.transcriptsDir) {
-    mkdirSync(args.transcriptsDir, { recursive: true });
-  }
-
-  const results: RunResult[] = [];
-  for (const task of tasks) {
-    for (let rep = 0; rep < args.repeats; rep++) {
-      for (const mode of args.modes) {
-        const db = cloneDb(task.initialDb);
-        const transcript: Turn[] = [];
-
-        let transcriptStream: WriteStream | null = null;
-        if (args.transcriptsDir) {
-          const fname = `${task.id}.${mode}.r${rep + 1}.jsonl`;
-          transcriptStream = openTranscriptFile(join(args.transcriptsDir, fname), {
-            version: 1,
-            source: `bench/${mode}`,
-            model: args.model,
-            task: task.id,
-            mode,
-            repeat: rep + 1,
-            startedAt: new Date().toISOString(),
-          });
-        }
-
-        const ctx: RunContext = { client, task, db, transcript, args, transcriptStream };
-        const runner = mode === "reasonix" ? runReasonix : runBaseline;
-        const started = Date.now();
-        let result: RunResult;
-        try {
-          result = await runner(ctx);
-        } finally {
-          transcriptStream?.end();
-        }
-        const elapsed = ((Date.now() - started) / 1000).toFixed(1);
-        console.log(
-          `[${task.id}/${mode}/r${rep + 1}] pass=${result.pass} turns=${result.turns} ` +
-            `cache=${(result.cacheHitRatio * 100).toFixed(1)}% cost=$${result.costUsd.toFixed(
-              6,
-            )} (${elapsed}s)`,
-        );
-        results.push(result);
-      }
-    }
-  }
-
-  const report: BenchReport = { meta: buildMeta(args, tasks.length), results };
-  writeReport(report, args.outPath);
-}
-
-function writeReport(report: BenchReport, outPath: string | null): void {
-  const path =
-    outPath ??
-    `benchmarks/tau-bench/results-${new Date().toISOString().replace(/[:.]/g, "-")}.json`;
-  writeFileSync(path, `${JSON.stringify(report, null, 2)}\n`, "utf8");
-  console.log(`wrote ${path}`);
-}
-
-function isMain(): boolean {
-  const entry = process.argv[1];
-  if (!entry) return false;
-  return import.meta.url === pathToFileURL(entry).href;
-}
-
-if (isMain()) {
-  main().catch((err) => {
-    console.error(err);
-    process.exit(1);
-  });
-}
-
-export { main as runBench };

benchmarks/tau-bench/tasks.ts

@@ -1,323 +0,0 @@
-/** Seed retail tasks — DB-end-state predicates avoid LLM-judge flakiness in the reproducibility report. */
-
-import { getRow, setField } from "./db.js";
-import type { TaskDefinition, ToolFactory, WorldState } from "./types.js";
-
-const RETAIL_SYSTEM = `You are a retail support agent. Use the tools to help the user.
-Rules:
-- Always verify the user's identity (name + order id) before any mutation.
-- Never invent order ids or emails.
-- If a request is outside your tools, say so honestly.
-- Be concise.`;
-
-function retailSeed(): WorldState {
-  return {
-    users: {
-      u_ari: { name: "Ari Chen", email: "[email protected]" },
-      u_bo: { name: "Bo Wang", email: "[email protected]" },
-      u_cai: { name: "Cai Lin", email: "[email protected]" },
-      u_dev: { name: "Dev Patel", email: "[email protected]" },
-    },
-    orders: {
-      o_1001: {
-        userId: "u_ari",
-        status: "shipped",
-        address: "1 Elm St, SF, CA 94110",
-        item: "wool sweater M",
-        price: 89.0,
-      },
-      o_1002: {
-        userId: "u_bo",
-        status: "processing",
-        address: "22 Oak Rd, NYC, NY 10001",
-        item: "running shoes 10",
-        price: 140.0,
-      },
-      o_1003: {
-        userId: "u_cai",
-        status: "delivered",
-        address: "9 Pine Ave, Austin, TX 78701",
-        item: "desk lamp",
-        price: 55.0,
-      },
-      o_1004: {
-        userId: "u_dev",
-        status: "processing",
-        address: "4 Maple Ln, Seattle, WA 98101",
-        item: "kettle",
-        price: 45.0,
-      },
-      o_1005: {
-        userId: "u_ari",
-        status: "delivered",
-        address: "1 Elm St, SF, CA 94110",
-        item: "notebook pack",
-        price: 22.0,
-      },
-    },
-    refunds: {},
-  };
-}
-
-const lookupOrder: ToolFactory = (db) => ({
-  name: "lookup_order",
-  description: "Look up an order by id. Returns { userId, status, address, item, price } or null.",
-  parameters: {
-    type: "object",
-    properties: { orderId: { type: "string" } },
-    required: ["orderId"],
-  },
-  fn: ({ orderId }: { orderId: string }) => {
-    const row = getRow(db, "orders", orderId);
-    return row ? { orderId, ...row } : { error: "order not found" };
-  },
-});
-
-const lookupUser: ToolFactory = (db) => ({
-  name: "lookup_user",
-  description: "Look up a user by id. Returns { name, email } or error.",
-  parameters: {
-    type: "object",
-    properties: { userId: { type: "string" } },
-    required: ["userId"],
-  },
-  fn: ({ userId }: { userId: string }) => {
-    const row = getRow(db, "users", userId);
-    return row ? { userId, ...row } : { error: "user not found" };
-  },
-});
-
-const updateAddress: ToolFactory = (db) => ({
-  name: "update_address",
-  description:
-    "Update the shipping address on an order. Only allowed if status is 'processing'. Returns ok/error.",
-  parameters: {
-    type: "object",
-    properties: {
-      orderId: { type: "string" },
-      address: { type: "string" },
-    },
-    required: ["orderId", "address"],
-  },
-  fn: ({ orderId, address }: { orderId: string; address: string }) => {
-    const row = getRow(db, "orders", orderId);
-    if (!row) return { error: "order not found" };
-    if (row.status !== "processing") return { error: `cannot edit: status=${row.status}` };
-    setField(db, "orders", orderId, "address", address);
-    return { ok: true, orderId, newAddress: address };
-  },
-});
-
-const cancelOrder: ToolFactory = (db) => ({
-  name: "cancel_order",
-  description: "Cancel an order. Only allowed if status is 'processing'. Returns ok/error.",
-  parameters: {
-    type: "object",
-    properties: { orderId: { type: "string" } },
-    required: ["orderId"],
-  },
-  fn: ({ orderId }: { orderId: string }) => {
-    const row = getRow(db, "orders", orderId);
-    if (!row) return { error: "order not found" };
-    if (row.status !== "processing") return { error: `cannot cancel: status=${row.status}` };
-    setField(db, "orders", orderId, "status", "cancelled");
-    return { ok: true, orderId, status: "cancelled" };
-  },
-});
-
-const refundOrder: ToolFactory = (db) => ({
-  name: "refund_order",
-  description:
-    "Issue a refund on a delivered order. Records an entry in refunds. Returns ok/error.",
-  parameters: {
-    type: "object",
-    properties: {
-      orderId: { type: "string" },
-      reason: { type: "string" },
-    },
-    required: ["orderId", "reason"],
-  },
-  fn: ({ orderId, reason }: { orderId: string; reason: string }) => {
-    const row = getRow(db, "orders", orderId);
-    if (!row) return { error: "order not found" };
-    if (row.status !== "delivered") return { error: `cannot refund: status=${row.status}` };
-    db.refunds[orderId] = { orderId, reason, amount: row.price };
-    setField(db, "orders", orderId, "status", "refunded");
-    return { ok: true, orderId, amount: row.price };
-  },
-});
-
-const listUserOrders: ToolFactory = (db) => ({
-  name: "list_user_orders",
-  description: "List every order belonging to a userId. Returns an array of orders.",
-  parameters: {
-    type: "object",
-    properties: { userId: { type: "string" } },
-    required: ["userId"],
-  },
-  fn: ({ userId }: { userId: string }) => {
-    const all = db.orders ?? {};
-    const out: unknown[] = [];
-    for (const [orderId, row] of Object.entries(all)) {
-      if (row.userId === userId) out.push({ orderId, ...row });
-    }
-    return out;
-  },
-});
-
-const ALL_TOOLS = [
-  lookupOrder,
-  lookupUser,
-  updateAddress,
-  cancelOrder,
-  refundOrder,
-  listUserOrders,
-];
-
-export const TASKS: TaskDefinition[] = [
-  {
-    id: "t01_address_happy",
-    description: "User wants to change the shipping address of a processing order.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "polite, brief",
-      goal: "Change the shipping address on order o_1002 to '5 Birch Rd, NYC, NY 10001'.",
-      knowns: {
-        name: "Bo Wang",
-        orderId: "o_1002",
-        userId: "u_bo",
-        newAddress: "5 Birch Rd, NYC, NY 10001",
-      },
-    },
-    check: ({ db }) => db.orders.o_1002?.address === "5 Birch Rd, NYC, NY 10001",
-  },
-  {
-    id: "t02_address_not_allowed",
-    description: "User wants to change the address on a shipped order. Agent should refuse.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "insistent",
-      goal: "Get the address on order o_1001 changed to '99 New St, SF, CA'.",
-      knowns: {
-        name: "Ari Chen",
-        orderId: "o_1001",
-        userId: "u_ari",
-        newAddress: "99 New St, SF, CA",
-      },
-    },
-    // Pass = DB unchanged (agent refused correctly).
-    check: ({ db }) => db.orders.o_1001?.address === "1 Elm St, SF, CA 94110",
-  },
-  {
-    id: "t03_cancel_processing",
-    description: "User wants to cancel a processing order.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "apologetic",
-      goal: "Cancel order o_1004.",
-      knowns: { name: "Dev Patel", orderId: "o_1004", userId: "u_dev" },
-    },
-    check: ({ db }) => db.orders.o_1004?.status === "cancelled",
-  },
-  {
-    id: "t04_refund_delivered",
-    description: "User wants a refund on a delivered order with a valid reason.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "unhappy but reasonable",
-      goal: "Get a refund on order o_1003 because the lamp arrived broken.",
-      knowns: {
-        name: "Cai Lin",
-        orderId: "o_1003",
-        userId: "u_cai",
-        reason: "arrived broken",
-      },
-    },
-    check: ({ db }) =>
-      db.orders.o_1003?.status === "refunded" && db.refunds.o_1003?.amount === 55.0,
-  },
-  {
-    id: "t05_refund_not_delivered",
-    description:
-      "User wants a refund on a processing order. Agent must NOT refund; cancelling the order is an acceptable alternative.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "demanding",
-      goal: "Get a refund on order o_1002 right now.",
-      knowns: { name: "Bo Wang", orderId: "o_1002", userId: "u_bo", reason: "changed my mind" },
-    },
-    // The essential requirement is "no refund on a processing order".
-    // Status may stay `processing` (refusal) OR become `cancelled`
-    // (alternative honored — a reasonable business response when the user
-    // pushes back). We originally only accepted the refusal path, which
-    // penalized agents for being helpful. Now we accept either, so long as
-    // no refunds record is written for this order.
-    check: ({ db }) => {
-      const status = db.orders.o_1002?.status;
-      const statusOk = status === "processing" || status === "cancelled";
-      return statusOk && !db.refunds.o_1002;
-    },
-  },
-  {
-    id: "t06_multi_order_lookup",
-    description: "User wants to hear their full order history, then cancels one.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "chatty",
-      goal: "First hear every order on file for user u_ari, then cancel any that are still processing.",
-      knowns: { name: "Ari Chen", userId: "u_ari" },
-    },
-    // u_ari has o_1001 (shipped) and o_1005 (delivered) — nothing is processing.
-    // Pass = both of Ari's orders unchanged.
-    check: ({ db }) =>
-      db.orders.o_1001?.status === "shipped" && db.orders.o_1005?.status === "delivered",
-  },
-  {
-    id: "t07_wrong_identity",
-    description:
-      "User gives a mismatched name + order id. Agent should not mutate on unverified identity.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "confident but wrong",
-      goal: "Claim to be 'Bo Wang' and ask to cancel order o_1004 (which actually belongs to Dev Patel).",
-      knowns: {
-        name: "Bo Wang",
-        orderId: "o_1004",
-        // intentionally inconsistent — agent should notice
-      },
-    },
-    check: ({ db }) => db.orders.o_1004?.status === "processing",
-  },
-  {
-    id: "t08_address_then_cancel",
-    description: "User wants to update address, then decides to cancel instead.",
-    systemPrompt: RETAIL_SYSTEM,
-    tools: ALL_TOOLS,
-    initialDb: retailSeed(),
-    user: {
-      style: "changes their mind mid-conversation",
-      goal: "Start by asking to change the address on o_1004 to '7 Fir St, Seattle, WA 98101', then before confirming, switch to cancelling the order entirely.",
-      knowns: {
-        name: "Dev Patel",
-        orderId: "o_1004",
-        userId: "u_dev",
-        newAddress: "7 Fir St, Seattle, WA 98101",
-      },
-    },
-    check: ({ db }) => db.orders.o_1004?.status === "cancelled",
-  },
-];

benchmarks/tau-bench/transcripts/README.md

@@ -1,57 +0,0 @@
-# Reference transcripts
-
-These are the raw `.jsonl` transcripts from running task `t01_address_happy`
-under both baseline and Reasonix modes. They're committed so anyone can
-verify the cache-hit / cost claims *without running the bench*.
-
-## Files
-
-| file | what it is |
-|---|---|
-| `t01_address_happy.baseline.r1.jsonl` | Naive cache-hostile agent's record of the run |
-| `t01_address_happy.reasonix.r1.jsonl` | Reasonix's record of the same task |
-| `t01_address_happy.diff.md` | Output of `reasonix diff` on the two above |
-| `mcp-demo.add.jsonl` | End-to-end run through the bundled demo MCP server. DeepSeek called the `add` tool; the second turn hit 96.6% cache, 94% cheaper than Claude at same token counts |
-| `mcp-filesystem.jsonl` | End-to-end run through the **official external** `@modelcontextprotocol/server-filesystem`. 5 turns, 4 tool calls including a permission-denied recovery. Overall cache 96.7%, 97% cheaper than Claude. Proof that Cache-First generalizes to third-party MCP servers without any code change on our side |
-| `mcp-multi-server.jsonl` | End-to-end run with **two MCP servers concurrently** — bundled demo (`demo_add`) + official `@modelcontextprotocol/server-filesystem` (`fs_write_file`). Model computed 17+25 on one server, wrote the result to a real file via the other. 5 turns, 4 tool calls across two subprocesses. **1 distinct prefix hash** held across all turns — Cache-First byte-stability survives running two MCP servers at once. Cache 81.1%, cost $0.00185, 95.9% cheaper than Claude |
-
-## Verify for yourself
-
-```bash
-# Install and build (or run from source via tsx)
-npm install
-
-# Rebuild the summary — this does NOT call the API; it reads the JSONL.
-npx reasonix replay benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r1.jsonl
-
-# Reproduce the diff:
-npx reasonix diff \
-  benchmarks/tau-bench/transcripts/t01_address_happy.baseline.r1.jsonl \
-  benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r1.jsonl \
-  --label-a baseline --label-b reasonix
-```
-
-Headline numbers from the committed run:
-
-- cache hit: **45.9% → 93.9% (+48.0pp)**
-- cost: **$0.001192 → $0.000953 (−20.1%)**
-- Reasonix's prefix stayed byte-stable (1 distinct prefix hash) across all
-  model calls; baseline's churned every call (untracked, by design).
-
-## Regenerate
-
-```bash
-export DEEPSEEK_API_KEY=sk-...
-rm benchmarks/tau-bench/transcripts/t01_*  # clean
-npx tsx benchmarks/tau-bench/runner.ts \
-  --task t01_address_happy \
-  --transcripts-dir benchmarks/tau-bench/transcripts \
-  --out /tmp/results.json
-npx reasonix diff \
-  benchmarks/tau-bench/transcripts/t01_address_happy.baseline.r1.jsonl \
-  benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r1.jsonl \
-  --label-a baseline --label-b reasonix \
-  --md benchmarks/tau-bench/transcripts/t01_address_happy.diff.md
-```
-
-Cost: ~$0.003 per full regeneration.

benchmarks/tau-bench/transcripts/mcp-demo.add.jsonl

@@ -1,6 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"reasonix run","model":"deepseek-chat","startedAt":"2026-04-22T00:50:25.468Z"}}
-{"ts":"2026-04-22T00:50:25.469Z","turn":1,"role":"user","content":"Use the add tool to compute 17+25, then briefly confirm the result."}
-{"ts":"2026-04-22T00:50:28.556Z","turn":1,"role":"assistant_final","content":"I'll compute 17 + 25 using the add tool.","usage":{"prompt_tokens":442,"completion_tokens":70,"total_tokens":512,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":442},"cost":0.00019634,"model":"deepseek-chat","prefixHash":"585d4a0a16fe84b7"}
-{"ts":"2026-04-22T00:50:28.557Z","turn":1,"role":"tool","content":"42","tool":"add","args":"{\"a\": 17, \"b\": 25}"}
-{"ts":"2026-04-22T00:50:29.858Z","turn":1,"role":"assistant_final","content":"The result is 42. 17 + 25 = 42.","usage":{"prompt_tokens":530,"completion_tokens":15,"total_tokens":545,"prompt_cache_hit_tokens":512,"prompt_cache_miss_tokens":18},"cost":0.0000572,"model":"deepseek-chat","prefixHash":"585d4a0a16fe84b7"}
-{"ts":"2026-04-22T00:50:29.858Z","turn":1,"role":"done","content":"The result is 42. 17 + 25 = 42."}

benchmarks/tau-bench/transcripts/mcp-filesystem.jsonl

@@ -1,24 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"reasonix run","model":"deepseek-chat","startedAt":"2026-04-22T00:57:35.368Z"}}
-{"ts":"2026-04-22T00:57:35.369Z","turn":1,"role":"user","content":"List the files here, then read README.md and summarize it in one sentence."}
-{"ts":"2026-04-22T00:57:38.133Z","turn":1,"role":"assistant_final","content":"I'll start by listing the files in the current directory, then read the README.md file and summarize it.","usage":{"prompt_tokens":2218,"completion_tokens":64,"total_tokens":2282,"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":1962},"cost":0.0006180599999999999,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T00:57:38.136Z","turn":1,"role":"tool","content":"ERROR: Access denied - path outside allowed directories: F:\\Reasonix not in C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"list_directory","args":"{\"path\": \".\"}"}
-{"ts":"2026-04-22T00:57:40.282Z","turn":1,"role":"assistant_final","content":"Let me first check what directories are allowed:","usage":{"prompt_tokens":2337,"completion_tokens":39,"total_tokens":2376,"prompt_cache_hit_tokens":2240,"prompt_cache_miss_tokens":97},"cost":0.00022589000000000002,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T00:57:40.285Z","turn":1,"role":"tool","content":"Allowed directories:\nC:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"list_allowed_directories","args":"{}"}
-{"ts":"2026-04-22T00:57:43.982Z","turn":1,"role":"assistant_final","content":"Now let me list the files in the allowed directory:","usage":{"prompt_tokens":2418,"completion_tokens":74,"total_tokens":2492,"prompt_cache_hit_tokens":2304,"prompt_cache_miss_tokens":114},"cost":0.00027346000000000006,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T00:57:43.983Z","turn":1,"role":"tool","content":"[FILE] a.txt\n[FILE] b.txt\n[FILE] README.md","tool":"list_directory","args":"{\"path\": \"C:\\\\Users\\\\yuhuahui\\\\AppData\\\\Local\\\\Temp\\\\mcp-fs-demo\"}"}
-{"ts":"2026-04-22T00:57:47.317Z","turn":1,"role":"assistant_final","content":"Now let me read the README.md file:","usage":{"prompt_tokens":2526,"completion_tokens":77,"total_tokens":2603,"prompt_cache_hit_tokens":2432,"prompt_cache_miss_tokens":94},"cost":0.00028031999999999997,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T00:57:47.322Z","turn":1,"role":"tool","content":"# Sample","tool":"read_text_file","args":"{\"path\": \"C:\\\\Users\\\\yuhuahui\\\\AppData\\\\Local\\\\Temp\\\\mcp-fs-demo\\\\README.md\"}"}
-{"ts":"2026-04-22T00:57:49.096Z","turn":1,"role":"assistant_final","content":"**Summary:** The directory contains three files (a.txt, b.txt, and README.md), and the README.md file simply contains the word \"Sample\".","usage":{"prompt_tokens":2622,"completion_tokens":33,"total_tokens":2655,"prompt_cache_hit_tokens":2560,"prompt_cache_miss_tokens":62},"cost":0.00023224000000000004,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T00:57:49.096Z","turn":1,"role":"done","content":"**Summary:** The directory contains three files (a.txt, b.txt, and README.md), and the README.md file simply contains the word \"Sample\"."}
-{"role":"_meta","meta":{"version":1,"source":"reasonix run","model":"deepseek-chat","startedAt":"2026-04-22T01:00:29.520Z"}}
-{"ts":"2026-04-22T01:00:29.521Z","turn":1,"role":"user","content":"List the files here, then read README.md and summarize it in one sentence."}
-{"ts":"2026-04-22T01:00:32.502Z","turn":1,"role":"assistant_final","content":"I'll start by listing the files in the current directory and then read the README.md file.","usage":{"prompt_tokens":2218,"completion_tokens":61,"total_tokens":2279,"prompt_cache_hit_tokens":2176,"prompt_cache_miss_tokens":42},"cost":0.00023076000000000005,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T01:00:32.505Z","turn":1,"role":"tool","content":"ERROR: Access denied - path outside allowed directories: F:\\Reasonix not in C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"list_directory","args":"{\"path\": \".\"}"}
-{"ts":"2026-04-22T01:00:34.833Z","turn":1,"role":"assistant_final","content":"Let me first check what directories are allowed:","usage":{"prompt_tokens":2334,"completion_tokens":39,"total_tokens":2373,"prompt_cache_hit_tokens":2240,"prompt_cache_miss_tokens":94},"cost":0.00022508,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T01:00:34.834Z","turn":1,"role":"tool","content":"Allowed directories:\nC:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"list_allowed_directories","args":"{}"}
-{"ts":"2026-04-22T01:00:38.039Z","turn":1,"role":"assistant_final","content":"Now let me list the files in the allowed directory:","usage":{"prompt_tokens":2415,"completion_tokens":74,"total_tokens":2489,"prompt_cache_hit_tokens":2304,"prompt_cache_miss_tokens":111},"cost":0.00027265000000000004,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T01:00:38.042Z","turn":1,"role":"tool","content":"[FILE] a.txt\n[FILE] b.txt\n[FILE] README.md","tool":"list_directory","args":"{\"path\": \"C:\\\\Users\\\\yuhuahui\\\\AppData\\\\Local\\\\Temp\\\\mcp-fs-demo\"}"}
-{"ts":"2026-04-22T01:00:41.597Z","turn":1,"role":"assistant_final","content":"Now I'll read the README.md file:","usage":{"prompt_tokens":2523,"completion_tokens":77,"total_tokens":2600,"prompt_cache_hit_tokens":2432,"prompt_cache_miss_tokens":91},"cost":0.00027951,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T01:00:41.601Z","turn":1,"role":"tool","content":"# Sample","tool":"read_text_file","args":"{\"path\": \"C:\\\\Users\\\\yuhuahui\\\\AppData\\\\Local\\\\Temp\\\\mcp-fs-demo\\\\README.md\"}"}
-{"ts":"2026-04-22T01:00:43.759Z","turn":1,"role":"assistant_final","content":"**Files in the directory:** a.txt, b.txt, README.md\n\n**README.md summary:** This is a sample README file.","usage":{"prompt_tokens":2619,"completion_tokens":29,"total_tokens":2648,"prompt_cache_hit_tokens":2560,"prompt_cache_miss_tokens":59},"cost":0.00022703000000000003,"model":"deepseek-chat","prefixHash":"6a7fff6a15fd5def"}
-{"ts":"2026-04-22T01:00:43.759Z","turn":1,"role":"done","content":"**Files in the directory:** a.txt, b.txt, README.md\n\n**README.md summary:** This is a sample README file."}

benchmarks/tau-bench/transcripts/mcp-multi-server.jsonl

@@ -1,12 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"reasonix run","model":"deepseek-chat","startedAt":"2026-04-22T01:39:54.789Z"}}
-{"ts":"2026-04-22T01:39:54.790Z","turn":1,"role":"user","content":"First use the add tool to compute 17+25, then use write_file to save the result to a file called answer.txt in the current working directory."}
-{"ts":"2026-04-22T01:39:58.843Z","turn":1,"role":"assistant_final","content":"I'll help you with that. First, I'll compute 17+25 using the add tool, then write the result to answer.txt.","usage":{"prompt_tokens":2404,"completion_tokens":87,"total_tokens":2491,"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":2148},"cost":0.0006935800000000001,"model":"deepseek-chat","prefixHash":"ea71560eab18d572"}
-{"ts":"2026-04-22T01:39:58.844Z","turn":1,"role":"tool","content":"42","tool":"demo_add","args":"{\"a\": 17, \"b\": 25}"}
-{"ts":"2026-04-22T01:40:02.898Z","turn":1,"role":"assistant_final","content":"Now I'll write the result (42) to a file called answer.txt in the current working directory:","usage":{"prompt_tokens":2509,"completion_tokens":81,"total_tokens":2590,"prompt_cache_hit_tokens":2432,"prompt_cache_miss_tokens":77},"cost":0.00028013,"model":"deepseek-chat","prefixHash":"ea71560eab18d572"}
-{"ts":"2026-04-22T01:40:02.900Z","turn":1,"role":"tool","content":"ERROR: Access denied - path outside allowed directories: F:\\Reasonix\\answer.txt not in C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"fs_write_file","args":"{\"path\": \"answer.txt\", \"content\": \"42\"}"}
-{"ts":"2026-04-22T01:40:05.263Z","turn":1,"role":"assistant_final","content":"I need to check which directories are allowed first. Let me check the allowed directories:","usage":{"prompt_tokens":2648,"completion_tokens":48,"total_tokens":2696,"prompt_cache_hit_tokens":2560,"prompt_cache_miss_tokens":88},"cost":0.00025576,"model":"deepseek-chat","prefixHash":"ea71560eab18d572"}
-{"ts":"2026-04-22T01:40:05.270Z","turn":1,"role":"tool","content":"Allowed directories:\nC:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo","tool":"fs_list_allowed_directories","args":"{}"}
-{"ts":"2026-04-22T01:40:09.422Z","turn":1,"role":"assistant_final","content":"Now I'll write the file to the allowed directory:","usage":{"prompt_tokens":2738,"completion_tokens":94,"total_tokens":2832,"prompt_cache_hit_tokens":2624,"prompt_cache_miss_tokens":114},"cost":0.00031786,"model":"deepseek-chat","prefixHash":"ea71560eab18d572"}
-{"ts":"2026-04-22T01:40:09.426Z","turn":1,"role":"tool","content":"Successfully wrote to C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo\\answer.txt","tool":"fs_write_file","args":"{\"path\": \"C:\\\\Users\\\\yuhuahui\\\\AppData\\\\Local\\\\Temp\\\\mcp-fs-demo\\\\answer.txt\", \"content\": \"42\"}"}
-{"ts":"2026-04-22T01:40:13.689Z","turn":1,"role":"assistant_final","content":"Perfect! I've completed both tasks:\n\n1. **Computed 17+25**: The result is 42\n2. **Saved to answer.txt**: Created the file `answer.txt` in the allowed directory with the content \"42\"\n\nThe file has been successfully written to `C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo\\answer.txt`.","usage":{"prompt_tokens":2878,"completion_tokens":83,"total_tokens":2961,"prompt_cache_hit_tokens":2816,"prompt_cache_miss_tokens":62},"cost":0.00030516,"model":"deepseek-chat","prefixHash":"ea71560eab18d572"}
-{"ts":"2026-04-22T01:40:13.689Z","turn":1,"role":"done","content":"Perfect! I've completed both tasks:\n\n1. **Computed 17+25**: The result is 42\n2. **Saved to answer.txt**: Created the file `answer.txt` in the allowed directory with the content \"42\"\n\nThe file has been successfully written to `C:\\Users\\yuhuahui\\AppData\\Local\\Temp\\mcp-fs-demo\\answer.txt`."}

benchmarks/tau-bench/transcripts/t01_address_happy.baseline.r1.jsonl

@@ -1,39 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":1,"startedAt":"2026-04-21T13:21:57.995Z"}}
-{"ts":"2026-04-21T13:22:00.112Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address for my order o_1002."}
-{"ts":"2026-04-21T13:22:03.142Z","turn":1,"role":"assistant_final","content":"I can help you update the shipping address for your order. First, I need to verify your identity. Could you please provide your name and order ID? You've already given me order ID o_1002, but I also need your name to proceed with the address update.","usage":{"prompt_tokens":733,"completion_tokens":56,"total_tokens":789,"prompt_cache_hit_tokens":64,"prompt_cache_miss_tokens":669},"cost":0.00024671000000000003,"model":"deepseek-chat"}
-{"ts":"2026-04-21T13:22:04.688Z","turn":2,"role":"user","content":"My name is Bo Wang."}
-{"ts":"2026-04-21T13:22:10.465Z","turn":2,"role":"assistant_final","content":"Thank you, Bo Wang. Let me first look up your order o_1002 to verify the details and check its current status.","usage":{"prompt_tokens":799,"completion_tokens":74,"total_tokens":873,"prompt_cache_hit_tokens":64,"prompt_cache_miss_tokens":735},"cost":0.00028433000000000006,"model":"deepseek-chat"}
-{"ts":"2026-04-21T13:22:10.465Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-21T13:22:10.465Z","turn":2,"role":"assistant_final","content":"Great! I can see your order o_1002 is currently in \"processing\" status, which means I can update the shipping address. What would you like the new shipping address to be?","usage":{"prompt_tokens":932,"completion_tokens":39,"total_tokens":971,"prompt_cache_hit_tokens":832,"prompt_cache_miss_tokens":100},"cost":0.00012814,"model":"deepseek-chat"}
-{"ts":"2026-04-21T13:22:11.830Z","turn":3,"role":"user","content":"Please change the shipping address to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-21T13:22:19.185Z","turn":3,"role":"assistant_final","content":"Perfect! I'll update the shipping address for your order o_1002 from \"22 Oak Rd, NYC, NY 10001\" to \"5 Birch Rd, NYC, NY 10001\".","usage":{"prompt_tokens":993,"completion_tokens":112,"total_tokens":1105,"prompt_cache_hit_tokens":64,"prompt_cache_miss_tokens":929},"cost":0.00037851,"model":"deepseek-chat"}
-{"ts":"2026-04-21T13:22:19.185Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-21T13:22:19.185Z","turn":3,"role":"assistant_final","content":"Success! I've updated the shipping address for your order o_1002 to \"5 Birch Rd, NYC, NY 10001\". Your order is still in processing status, and the new address will be used for shipping. Is there anything else I can help you with?","usage":{"prompt_tokens":1149,"completion_tokens":56,"total_tokens":1205,"prompt_cache_hit_tokens":1088,"prompt_cache_miss_tokens":61},"cost":0.00015423000000000002,"model":"deepseek-chat"}
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":1,"startedAt":"2026-04-29T16:07:39.762Z"}}
-{"ts":"2026-04-29T16:07:41.059Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:07:42.123Z","turn":1,"role":"assistant_final","content":"I'd be happy to help you update the shipping address on order o_1002. First, I need to verify your identity. Could you please tell me your name so I can look into this?","usage":{"prompt_tokens":715,"completion_tokens":42,"total_tokens":757,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.00011106100000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:43.266Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang."}
-{"ts":"2026-04-29T16:07:46.627Z","turn":2,"role":"assistant_final","content":"Thanks! Let me look up the order details to verify your identity and check the current status.","usage":{"prompt_tokens":768,"completion_tokens":67,"total_tokens":835,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":768},"cost":0.00012537800000000002,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:46.627Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:07:46.627Z","turn":2,"role":"assistant_final","content":"Good, the order is in \"processing\" status, so we can update the address. Let me also verify your identity.\n\nCould you please confirm your name or provide your user ID so I can match it with the order? The order belongs to user \"u_bo\". What's your name?","usage":{"prompt_tokens":888,"completion_tokens":60,"total_tokens":948,"prompt_cache_hit_tokens":640,"prompt_cache_miss_tokens":248},"cost":0.000069072,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:47.358Z","turn":3,"role":"user","content":"My name is Bo Wang, and my user ID is u_bo."}
-{"ts":"2026-04-29T16:07:49.861Z","turn":3,"role":"assistant_final","content":"Let me look up the user details to confirm.","usage":{"prompt_tokens":967,"completion_tokens":56,"total_tokens":1023,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":967},"cost":0.000149981,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:49.861Z","turn":3,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:07:49.861Z","turn":3,"role":"assistant_final","content":"Great, you're verified! The order o_1002 (running shoes, $140) is currently in \"processing\" status and is set to ship to **22 Oak Rd, NYC, NY 10001**.\n\nWhat new address would you like to update it to?","usage":{"prompt_tokens":1053,"completion_tokens":55,"total_tokens":1108,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":157},"cost":0.000062201,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:50.752Z","turn":4,"role":"user","content":"Please change the shipping address to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:07:53.488Z","turn":4,"role":"assistant_final","content":"","usage":{"prompt_tokens":1130,"completion_tokens":72,"total_tokens":1202,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1130},"cost":0.00017708600000000002,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:07:53.488Z","turn":4,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:07:53.488Z","turn":4,"role":"assistant_final","content":"The shipping address for order o_1002 has been updated successfully to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?","usage":{"prompt_tokens":1241,"completion_tokens":36,"total_tokens":1277,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":89},"cost":0.00005463500000000001,"model":"deepseek-chat"}
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":1,"startedAt":"2026-04-29T16:19:42.654Z"}}
-{"ts":"2026-04-29T16:19:43.504Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:19:44.637Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! First, I need to verify your identity. Could you please tell me your name and order ID so I can look into it?","usage":{"prompt_tokens":715,"completion_tokens":34,"total_tokens":749,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.000108837,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:19:45.518Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang and the order ID is o_1002."}
-{"ts":"2026-04-29T16:19:49.701Z","turn":2,"role":"assistant_final","content":"Let me look up the order and verify your details.","usage":{"prompt_tokens":769,"completion_tokens":59,"total_tokens":828,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":769},"cost":0.00012329300000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:19:49.701Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:19:49.701Z","turn":2,"role":"assistant_final","content":"The order is found. Now let me look up the user to verify your identity.","usage":{"prompt_tokens":881,"completion_tokens":62,"total_tokens":943,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":113},"cost":0.000054447,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:19:49.701Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:19:49.701Z","turn":2,"role":"assistant_final","content":"Great! I've confirmed you are **Bo Wang** and the order **o_1002** is yours. \n\nHere are the current details:\n- **Item:** Running shoes 10\n- **Status:** Processing ✅ (address can be updated)\n- **Current Address:** 22 Oak Rd, NYC, NY 10001\n\nWhat new address would you like to change it to?","usage":{"prompt_tokens":974,"completion_tokens":79,"total_tokens":1053,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":78},"cost":0.000057892,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:19:50.408Z","turn":3,"role":"user","content":"I'd like to change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:19:52.977Z","turn":3,"role":"assistant_final","content":"Let me update the address for you.","usage":{"prompt_tokens":1076,"completion_tokens":80,"total_tokens":1156,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1076},"cost":0.00017180400000000004,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:19:52.977Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:19:52.977Z","turn":3,"role":"assistant_final","content":"Done! The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help with?","usage":{"prompt_tokens":1194,"completion_tokens":38,"total_tokens":1232,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":170},"cost":0.000062866,"model":"deepseek-chat"}

benchmarks/tau-bench/transcripts/t01_address_happy.baseline.r2.jsonl

@@ -1,26 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":2,"startedAt":"2026-04-29T16:07:57.317Z"}}
-{"ts":"2026-04-29T16:07:58.462Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:07:59.695Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! First, I need to verify your identity. Could you please tell me your name so I can look into your order?","usage":{"prompt_tokens":715,"completion_tokens":32,"total_tokens":747,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.000108281,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:00.326Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang."}
-{"ts":"2026-04-29T16:08:04.410Z","turn":2,"role":"assistant_final","content":"Let me look up the order to check the details.","usage":{"prompt_tokens":758,"completion_tokens":59,"total_tokens":817,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":758},"cost":0.00012176400000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:04.410Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:04.410Z","turn":2,"role":"assistant_final","content":"Let me also look up your user info to confirm your identity.","usage":{"prompt_tokens":870,"completion_tokens":58,"total_tokens":928,"prompt_cache_hit_tokens":640,"prompt_cache_miss_tokens":230},"cost":0.00006601400000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:04.410Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:08:04.411Z","turn":2,"role":"assistant_final","content":"Great, I've confirmed you are **Bo Wang** with order **o_1002**. The order is currently **processing** and the current address is:\n\n> 22 Oak Rd, NYC, NY 10001\n\nWhat would you like the new shipping address to be?","usage":{"prompt_tokens":959,"completion_tokens":56,"total_tokens":1015,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":63},"cost":0.000049412999999999996,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:05.185Z","turn":3,"role":"user","content":"I'd like to change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:08:08.194Z","turn":3,"role":"assistant_final","content":"Let me update the shipping address for you.","usage":{"prompt_tokens":1038,"completion_tokens":81,"total_tokens":1119,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1038},"cost":0.00016680000000000002,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:08.194Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:08:08.194Z","turn":3,"role":"assistant_final","content":"Done! The shipping address for order **o_1002** has been updated to:\n\n> **5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?","usage":{"prompt_tokens":1157,"completion_tokens":41,"total_tokens":1198,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":133},"cost":0.00005855700000000001,"model":"deepseek-chat"}
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":2,"startedAt":"2026-04-29T16:20:05.321Z"}}
-{"ts":"2026-04-29T16:20:06.518Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:20:07.685Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! First, I need to verify your identity. Could you please tell me your name so I can look up your details?","usage":{"prompt_tokens":715,"completion_tokens":32,"total_tokens":747,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.000108281,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:08.298Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang."}
-{"ts":"2026-04-29T16:20:12.873Z","turn":2,"role":"assistant_final","content":"Let me look up your information and the order details.","usage":{"prompt_tokens":758,"completion_tokens":59,"total_tokens":817,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":758},"cost":0.00012176400000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:12.873Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:12.873Z","turn":2,"role":"assistant_final","content":"Let me also look up your user details.","usage":{"prompt_tokens":870,"completion_tokens":54,"total_tokens":924,"prompt_cache_hit_tokens":640,"prompt_cache_miss_tokens":230},"cost":0.000064902,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:12.873Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:12.873Z","turn":2,"role":"assistant_final","content":"I've confirmed you are **Bo Wang** and order **o_1002** is yours. Here are the current details:\n\n- **Item:** running shoes 10\n- **Current Address:** 22 Oak Rd, NYC, NY 10001\n- **Status:** Processing ✅ (eligible for address update)\n\nWhat would you like the new shipping address to be?","usage":{"prompt_tokens":955,"completion_tokens":75,"total_tokens":1030,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":59},"cost":0.000054139000000000005,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:13.791Z","turn":3,"role":"user","content":"Please change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:20:16.480Z","turn":3,"role":"assistant_final","content":"Let me update the address for you.","usage":{"prompt_tokens":1050,"completion_tokens":80,"total_tokens":1130,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1050},"cost":0.00016819000000000003,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:16.480Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:16.480Z","turn":3,"role":"assistant_final","content":"Done! The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?","usage":{"prompt_tokens":1168,"completion_tokens":39,"total_tokens":1207,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":144},"cost":0.00005953,"model":"deepseek-chat"}

benchmarks/tau-bench/transcripts/t01_address_happy.baseline.r3.jsonl

@@ -1,24 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":3,"startedAt":"2026-04-29T16:08:11.479Z"}}
-{"ts":"2026-04-29T16:08:12.282Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:08:13.202Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! First, I need to verify your identity. Could you please tell me your name and user ID so I can look it up?","usage":{"prompt_tokens":715,"completion_tokens":34,"total_tokens":749,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.000108837,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:14.024Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang and my user ID is u_bo."}
-{"ts":"2026-04-29T16:08:17.451Z","turn":2,"role":"assistant_final","content":"Let me look up your info and check the order details.","usage":{"prompt_tokens":768,"completion_tokens":93,"total_tokens":861,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":768},"cost":0.000132606,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:17.451Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:08:17.451Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:17.451Z","turn":2,"role":"assistant_final","content":"Great, I've confirmed you're **Bo Wang** (u_bo) and order **o_1002** belongs to you. The order is currently **processing** and the current shipping address is:\n\n**22 Oak Rd, NYC, NY 10001**\n\nWhat new address would you like to update it to?","usage":{"prompt_tokens":941,"completion_tokens":65,"total_tokens":1006,"prompt_cache_hit_tokens":640,"prompt_cache_miss_tokens":301},"cost":0.000077829,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:18.366Z","turn":3,"role":"user","content":"Please change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:08:21.602Z","turn":3,"role":"assistant_final","content":"Let me update the address now.","usage":{"prompt_tokens":1026,"completion_tokens":79,"total_tokens":1105,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1026},"cost":0.00016457600000000002,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:08:21.602Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:08:21.602Z","turn":3,"role":"assistant_final","content":"Your shipping address has been updated! Here's the summary:\n\n- **Order:** o_1002 (running shoes - $140)\n- **Old address:** 22 Oak Rd, NYC, NY 10001\n- **New address:** 5 Birch Rd, NYC, NY 10001\n\nIs there anything else I can help you with?","usage":{"prompt_tokens":1143,"completion_tokens":71,"total_tokens":1214,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":119},"cost":0.00006495100000000001,"model":"deepseek-chat"}
-{"role":"_meta","meta":{"version":1,"source":"bench/baseline","model":"deepseek-chat","task":"t01_address_happy","mode":"baseline","repeat":3,"startedAt":"2026-04-29T16:20:30.878Z"}}
-{"ts":"2026-04-29T16:20:31.865Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:20:32.841Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! First, I need to verify your identity. Could you please tell me your name and user ID so I can look into your order?","usage":{"prompt_tokens":715,"completion_tokens":35,"total_tokens":750,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":715},"cost":0.00010911500000000001,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:33.559Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang and my user ID is u_bo."}
-{"ts":"2026-04-29T16:20:36.980Z","turn":2,"role":"assistant_final","content":"Let me look up your information and the order details.","usage":{"prompt_tokens":769,"completion_tokens":92,"total_tokens":861,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":769},"cost":0.00013246700000000002,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:36.980Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:36.980Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:36.980Z","turn":2,"role":"assistant_final","content":"Great, I've confirmed you're **Bo Wang** (user: u_bo), and order **o_1002** (running shoes, $140) is currently in **processing** status with the address:\n\n> **22 Oak Rd, NYC, NY 10001**\n\nSince the order is still processing, you're eligible to update the address. What would you like the new shipping address to be?","usage":{"prompt_tokens":941,"completion_tokens":83,"total_tokens":1024,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":173},"cost":0.000068625,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:37.943Z","turn":3,"role":"user","content":"I'd like to change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:20:40.543Z","turn":3,"role":"assistant_final","content":"Let me update that for you.","usage":{"prompt_tokens":1047,"completion_tokens":79,"total_tokens":1126,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":1047},"cost":0.000167495,"model":"deepseek-chat"}
-{"ts":"2026-04-29T16:20:40.543Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:40.543Z","turn":3,"role":"assistant_final","content":"All done! The shipping address for order **o_1002** has been updated to:\n\n> **5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?","usage":{"prompt_tokens":1164,"completion_tokens":42,"total_tokens":1206,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":140},"cost":0.00005980800000000001,"model":"deepseek-chat"}

benchmarks/tau-bench/transcripts/t01_address_happy.diff.md

@@ -1,45 +0,0 @@
-# Transcript diff: baseline vs reasonix
-
-## Meta
-
-| | baseline | reasonix |
-|---|---|---|
-| source | bench/baseline | bench/reasonix |
-| model | deepseek-chat | deepseek-chat |
-| task | t01_address_happy | t01_address_happy |
-| startedAt | 2026-04-21T13:21:57.995Z | 2026-04-21T13:22:21.292Z |
-
-## Summary
-
-| metric | baseline | reasonix | delta |
-|---|---:|---:|---:|
-| model calls | 5 | 6 | +1 |
-| user turns | 3 | 3 | 0 |
-| tool calls | 2 | 3 | +1 |
-| cache hit | 45.9% | 93.9% | **+48.0pp** |
-| cost (USD) | $0.001192 | $0.000953 | -20.1% |
-| prefix hashes | 0 | 1 | — |
-
-## Turn-by-turn
-
-| turn | kind | baseline tool calls | reasonix tool calls | note |
-|---:|:---:|---|---|---|
-| 1 | diverge | — | lookup_order | tool calls differ: A=[—] B=[lookup_order] |
-| 2 | diverge | lookup_order | lookup_user | tool calls differ: A=[lookup_order] B=[lookup_user] |
-| 3 | diverge | update_address | update_address | text similarity 68% |
-
-## First divergence (turn 1)
-
-tool calls differ: A=[—] B=[lookup_order]
-
-**baseline:**
-
-```
-I can help you update the shipping address for your order. First, I need to verify your identity. Could you please provide your name and order ID? You've already given me order ID o_1002, but I also need your name to proceed with the address update.
-```
-
-**reasonix:**
-
-```
-Great! I can see order o_1002 is in "processing" status, which means I can update the shipping address. However, I still need to verify your identity. Could you please provide your name so I can look up your user information using the user ID associated with this order?
-```

benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r1.jsonl

@@ -1,42 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":1,"startedAt":"2026-04-21T13:22:21.292Z"}}
-{"ts":"2026-04-21T13:22:22.687Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address for my order o_1002."}
-{"ts":"2026-04-21T13:22:27.201Z","turn":1,"role":"assistant_final","content":"I can help you update the shipping address for your order o_1002. First, I need to verify your identity. Could you please provide your name and order ID? I'll look up your order to check its current status and verify the details.","usage":{"prompt_tokens":713,"completion_tokens":98,"total_tokens":811,"prompt_cache_hit_tokens":704,"prompt_cache_miss_tokens":9},"cost":0.00015951000000000002,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:27.201Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-21T13:22:30.465Z","turn":1,"role":"assistant_final","content":"Great! I can see order o_1002 is in \"processing\" status, which means I can update the shipping address. However, I still need to verify your identity. Could you please provide your name so I can look up your user information using the user ID associated with this order?","usage":{"prompt_tokens":870,"completion_tokens":59,"total_tokens":929,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":102},"cost":0.00014620000000000003,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:30.465Z","turn":1,"role":"done","content":"Great! I can see order o_1002 is in \"processing\" status, which means I can update the shipping address. However, I still need to verify your identity. Could you please provide your name so I can look up your user information using the user ID associated with this order?"}
-{"ts":"2026-04-21T13:22:31.801Z","turn":2,"role":"user","content":"My name is Bo Wang."}
-{"ts":"2026-04-21T13:22:34.932Z","turn":2,"role":"assistant_final","content":"Perfect, thank you Bo Wang. Let me verify your identity by looking up your user information.","usage":{"prompt_tokens":939,"completion_tokens":64,"total_tokens":1003,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":43},"cost":0.00014473,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:34.932Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-21T13:22:38.670Z","turn":2,"role":"assistant_final","content":"Excellent! Your identity has been verified, Bo Wang. I can see your order o_1002 is currently in \"processing\" status with the shipping address \"22 Oak Rd, NYC, NY 10001\".\n\nWhat would you like the new shipping address to be? Please provide the complete new address so I can update it for you.","usage":{"prompt_tokens":1039,"completion_tokens":68,"total_tokens":1107,"prompt_cache_hit_tokens":960,"prompt_cache_miss_tokens":79},"cost":0.00016333000000000002,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:38.670Z","turn":2,"role":"done","content":"Excellent! Your identity has been verified, Bo Wang. I can see your order o_1002 is currently in \"processing\" status with the shipping address \"22 Oak Rd, NYC, NY 10001\".\n\nWhat would you like the new shipping address to be? Please provide the complete new address so I can update it for you."}
-{"ts":"2026-04-21T13:22:41.867Z","turn":3,"role":"user","content":"Please change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-21T13:22:45.479Z","turn":3,"role":"assistant_final","content":"","usage":{"prompt_tokens":1127,"completion_tokens":72,"total_tokens":1199,"prompt_cache_hit_tokens":1088,"prompt_cache_miss_tokens":39},"cost":0.00016589000000000003,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:45.479Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-21T13:22:48.394Z","turn":3,"role":"assistant_final","content":"Perfect! I've successfully updated the shipping address for your order o_1002. The new address is now \"5 Birch Rd, NYC, NY 10001\". \n\nYour order is still in processing status and will be shipped to this new address. Is there anything else I can help you with today?","usage":{"prompt_tokens":1243,"completion_tokens":62,"total_tokens":1305,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":91},"cost":0.00017341000000000004,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-21T13:22:48.394Z","turn":3,"role":"done","content":"Perfect! I've successfully updated the shipping address for your order o_1002. The new address is now \"5 Birch Rd, NYC, NY 10001\". \n\nYour order is still in processing status and will be shipped to this new address. Is there anything else I can help you with today?"}
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":1,"startedAt":"2026-04-29T16:07:54.314Z"}}
-{"ts":"2026-04-29T16:07:55.377Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:07:57.257Z","turn":1,"role":"assistant_final","content":"Sure! Let me start by looking up your order.","usage":{"prompt_tokens":774,"completion_tokens":98,"total_tokens":872,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":774},"cost":0.00013483,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:07:57.258Z","turn":1,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:07:57.258Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:07:57.258Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:07:57.317Z","turn":1,"role":"error","content":"","error":"Bad request (DeepSeek 400): The `reasoning_content` in the thinking mode must be passed back to the API."}
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":1,"startedAt":"2026-04-29T16:19:53.726Z"}}
-{"ts":"2026-04-29T16:19:54.663Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:19:56.901Z","turn":1,"role":"assistant_final","content":"Hello! I'd be happy to help you with updating your shipping address. First, I need to verify your identity. Let me look up the order details.","usage":{"prompt_tokens":774,"completion_tokens":113,"total_tokens":887,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":6},"cost":0.000053752,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:19:56.901Z","turn":1,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:19:56.901Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:19:56.901Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:19:58.403Z","turn":1,"role":"assistant_final","content":"I can see the order belongs to user ID \"u_bo\". Could you please tell me your name so I can verify your identity? Also, could you confirm what you'd like the new shipping address to be?","usage":{"prompt_tokens":941,"completion_tokens":66,"total_tokens":1007,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":173},"cost":0.000063899,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:19:58.403Z","turn":1,"role":"done","content":"I can see the order belongs to user ID \"u_bo\". Could you please tell me your name so I can verify your identity? Also, could you confirm what you'd like the new shipping address to be?"}
-{"ts":"2026-04-29T16:19:59.485Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang. I’d like the shipping address changed to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:20:00.956Z","turn":2,"role":"assistant_final","content":"Let me verify your identity.","usage":{"prompt_tokens":1040,"completion_tokens":69,"total_tokens":1109,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":144},"cost":0.000064286,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:00.956Z","turn":2,"role":"tool_start","content":"","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:00.956Z","turn":2,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:00.956Z","turn":2,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:03.176Z","turn":2,"role":"assistant_final","content":"Great, you're verified! The order is currently in \"processing\" status, so I can update the address. Let me do that now.","usage":{"prompt_tokens":1140,"completion_tokens":126,"total_tokens":1266,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":116},"cost":0.00007982400000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:03.176Z","turn":2,"role":"tool_start","content":"","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:03.176Z","turn":2,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:03.176Z","turn":2,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:04.374Z","turn":2,"role":"assistant_final","content":"Done! Your shipping address for order **o_1002** has been updated to:\n\n**5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?","usage":{"prompt_tokens":1305,"completion_tokens":48,"total_tokens":1353,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":153},"cost":0.00006686700000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:04.374Z","turn":2,"role":"done","content":"Done! Your shipping address for order **o_1002** has been updated to:\n\n**5 Birch Rd, NYC, NY 10001**\n\nIs there anything else I can help you with?"}

benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r2.jsonl

@@ -1,29 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":2,"startedAt":"2026-04-29T16:08:08.866Z"}}
-{"ts":"2026-04-29T16:08:09.732Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:08:11.405Z","turn":1,"role":"assistant_final","content":"Sure, let me look up your order first.","usage":{"prompt_tokens":774,"completion_tokens":74,"total_tokens":848,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":6},"cost":0.00004291,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:08:11.405Z","turn":1,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:11.405Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:11.405Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:08:11.479Z","turn":1,"role":"error","content":"","error":"Bad request (DeepSeek 400): The `reasoning_content` in the thinking mode must be passed back to the API."}
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":2,"startedAt":"2026-04-29T16:20:17.219Z"}}
-{"ts":"2026-04-29T16:20:18.111Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:20:19.984Z","turn":1,"role":"assistant_final","content":"Let me start by looking up the order details.","usage":{"prompt_tokens":774,"completion_tokens":96,"total_tokens":870,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":6},"cost":0.000049026,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:19.984Z","turn":1,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:19.984Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:19.984Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:21.800Z","turn":1,"role":"assistant_final","content":"I can see the order belongs to user \"u_bo\". Let me look up your account.","usage":{"prompt_tokens":924,"completion_tokens":87,"total_tokens":1011,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":156},"cost":0.00006737400000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:21.800Z","turn":1,"role":"tool_start","content":"","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:21.801Z","turn":1,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:21.801Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:24.028Z","turn":1,"role":"assistant_final","content":"I see this order belongs to **Bo Wang** (email: [email protected]). Could you please confirm your name (first and last) and your order ID so I can verify your identity before making any changes?","usage":{"prompt_tokens":1042,"completion_tokens":102,"total_tokens":1144,"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":146},"cost":0.000073738,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:24.028Z","turn":1,"role":"done","content":"I see this order belongs to **Bo Wang** (email: [email protected]). Could you please confirm your name (first and last) and your order ID so I can verify your identity before making any changes?"}
-{"ts":"2026-04-29T16:20:24.971Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang and the order ID is o_1002."}
-{"ts":"2026-04-29T16:20:26.698Z","turn":2,"role":"assistant_final","content":"Great, your identity is verified! The order status is **processing**, so I can update the address. What would you like the new shipping address to be?","usage":{"prompt_tokens":1165,"completion_tokens":78,"total_tokens":1243,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":141},"cost":0.000069955,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:26.698Z","turn":2,"role":"done","content":"Great, your identity is verified! The order status is **processing**, so I can update the address. What would you like the new shipping address to be?"}
-{"ts":"2026-04-29T16:20:27.502Z","turn":3,"role":"user","content":"Please change the shipping address to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:20:29.108Z","turn":3,"role":"assistant_final","content":"","usage":{"prompt_tokens":1219,"completion_tokens":86,"total_tokens":1305,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":67},"cost":0.00006547700000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:29.108Z","turn":3,"role":"tool_start","content":"","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:29.108Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:29.108Z","turn":3,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:30.143Z","turn":3,"role":"assistant_final","content":"The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?","usage":{"prompt_tokens":1344,"completion_tokens":44,"total_tokens":1388,"prompt_cache_hit_tokens":1280,"prompt_cache_miss_tokens":64},"cost":0.000056968,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:30.143Z","turn":3,"role":"done","content":"The shipping address for order **o_1002** has been updated to **5 Birch Rd, NYC, NY 10001**. Is there anything else I can help you with?"}

benchmarks/tau-bench/transcripts/t01_address_happy.reasonix.r3.jsonl

@@ -1,32 +0,0 @@
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":3,"startedAt":"2026-04-29T16:08:22.300Z"}}
-{"ts":"2026-04-29T16:08:23.242Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:08:24.786Z","turn":1,"role":"assistant_final","content":"I'd be happy to help you update the shipping address for order o_1002. First, I need to verify your identity. Could you please tell me your name and any other identifying information so I can look up the order?","usage":{"prompt_tokens":774,"completion_tokens":79,"total_tokens":853,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":6},"cost":0.000044300000000000006,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:08:24.786Z","turn":1,"role":"done","content":"I'd be happy to help you update the shipping address for order o_1002. First, I need to verify your identity. Could you please tell me your name and any other identifying information so I can look up the order?"}
-{"ts":"2026-04-29T16:08:26.037Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang."}
-{"ts":"2026-04-29T16:08:27.678Z","turn":2,"role":"assistant_final","content":"","usage":{"prompt_tokens":833,"completion_tokens":67,"total_tokens":900,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":65},"cost":0.00004916500000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:08:27.678Z","turn":2,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:27.679Z","turn":2,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:08:27.679Z","turn":2,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:08:27.736Z","turn":2,"role":"error","content":"","error":"Bad request (DeepSeek 400): The `reasoning_content` in the thinking mode must be passed back to the API."}
-{"role":"_meta","meta":{"version":1,"source":"bench/reasonix","model":"deepseek-chat","task":"t01_address_happy","mode":"reasonix","repeat":3,"startedAt":"2026-04-29T16:20:41.502Z"}}
-{"ts":"2026-04-29T16:20:42.785Z","turn":1,"role":"user","content":"Hi, I need to update the shipping address on my order o_1002. Could you help me with that?"}
-{"ts":"2026-04-29T16:20:44.758Z","turn":1,"role":"assistant_final","content":"I'd be happy to help! Let me first look up your order details.","usage":{"prompt_tokens":774,"completion_tokens":94,"total_tokens":868,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":6},"cost":0.00004847,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:44.758Z","turn":1,"role":"tool_start","content":"","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:44.758Z","turn":1,"role":"tool","content":"{\"orderId\":\"o_1002\",\"userId\":\"u_bo\",\"status\":\"processing\",\"address\":\"22 Oak Rd, NYC, NY 10001\",\"item\":\"running shoes 10\",\"price\":140}","tool":"lookup_order","args":"{\"orderId\": \"o_1002\"}"}
-{"ts":"2026-04-29T16:20:44.758Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:46.765Z","turn":1,"role":"assistant_final","content":"Let me check the user info to verify your identity.","usage":{"prompt_tokens":922,"completion_tokens":110,"total_tokens":1032,"prompt_cache_hit_tokens":768,"prompt_cache_miss_tokens":154},"cost":0.00007349,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:46.765Z","turn":1,"role":"tool_start","content":"","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:46.765Z","turn":1,"role":"tool","content":"{\"userId\":\"u_bo\",\"name\":\"Bo Wang\",\"email\":\"[email protected]\"}","tool":"lookup_user","args":"{\"userId\": \"u_bo\"}"}
-{"ts":"2026-04-29T16:20:46.765Z","turn":1,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:48.272Z","turn":1,"role":"assistant_final","content":"I can see the order belongs to **Bo Wang**. For security, could you please confirm your name and order ID so I can proceed with the address update?","usage":{"prompt_tokens":1063,"completion_tokens":75,"total_tokens":1138,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":39},"cost":0.000054943000000000006,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:48.272Z","turn":1,"role":"done","content":"I can see the order belongs to **Bo Wang**. For security, could you please confirm your name and order ID so I can proceed with the address update?"}
-{"ts":"2026-04-29T16:20:49.080Z","turn":2,"role":"user","content":"Sure, my name is Bo Wang and the order ID is o_1002."}
-{"ts":"2026-04-29T16:20:50.348Z","turn":2,"role":"assistant_final","content":"Great, identity verified! What would you like the new shipping address to be?","usage":{"prompt_tokens":1159,"completion_tokens":47,"total_tokens":1206,"prompt_cache_hit_tokens":1024,"prompt_cache_miss_tokens":135},"cost":0.000060503,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:50.348Z","turn":2,"role":"done","content":"Great, identity verified! What would you like the new shipping address to be?"}
-{"ts":"2026-04-29T16:20:51.052Z","turn":3,"role":"user","content":"Please change it to 5 Birch Rd, NYC, NY 10001."}
-{"ts":"2026-04-29T16:20:52.906Z","turn":3,"role":"assistant_final","content":"","usage":{"prompt_tokens":1195,"completion_tokens":93,"total_tokens":1288,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":43},"cost":0.000064087,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:52.906Z","turn":3,"role":"tool_start","content":"","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:52.906Z","turn":3,"role":"tool","content":"{\"ok\":true,\"orderId\":\"o_1002\",\"newAddress\":\"5 Birch Rd, NYC, NY 10001\"}","tool":"update_address","args":"{\"orderId\": \"o_1002\", \"address\": \"5 Birch Rd, NYC, NY 10001\"}"}
-{"ts":"2026-04-29T16:20:52.906Z","turn":3,"role":"status","content":"tool result uploaded · model thinking before next response…"}
-{"ts":"2026-04-29T16:20:54.781Z","turn":3,"role":"assistant_final","content":"Your shipping address has been updated successfully! Here's a summary:\n\n- **Order ID:** o_1002\n- **Item:** running shoes 10\n- **Old Address:** 22 Oak Rd, NYC, NY 10001\n- **New Address:** 5 Birch Rd, NYC, NY 10001\n\nIs there anything else I can help you with?","usage":{"prompt_tokens":1327,"completion_tokens":84,"total_tokens":1411,"prompt_cache_hit_tokens":1152,"prompt_cache_miss_tokens":175},"cost":0.00007993300000000001,"model":"deepseek-chat","prefixHash":"fcb40cfd555bf805"}
-{"ts":"2026-04-29T16:20:54.781Z","turn":3,"role":"done","content":"Your shipping address has been updated successfully! Here's a summary:\n\n- **Order ID:** o_1002\n- **Item:** running shoes 10\n- **Old Address:** 22 Oak Rd, NYC, NY 10001\n- **New Address:** 5 Birch Rd, NYC, NY 10001\n\nIs there anything else I can help you with?"}