336 Commits
Author SHA1 Message Date
BillyOutlast 0aacd68f0d Merge pull request #317 from Heretek-AI/fixup/test-env-guard
fix(tooling): remove ambient web-provider settings before any suite loads (Refs #312)
2026-10-11 16:36:08 -04:00
John SmithandClaude Opus 5.5 758b22c58e fix(tooling): remove ambient web-provider settings before any suite loads (Refs #312)
#314 scrubbed BRAVE_API_KEY per test file, which leaves the next research
test free to reintroduce the leak. A bun test preload now deletes every
setting the research providers and the gateway backend read, for every run
started from the repo root. Tests that need one still set their own value.

- scripts/test-preload.ts: the scrub.
- bunfig.toml: [test] preload.
- core/test/clean-room.test.ts: a probe suite started with all six settings
  exported must see none of them.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-11 16:28:44 -04:00
BillyOutlast 9ec831939d Merge pull request #316 from Heretek-AI/fixup/release-2.3.0-notes
docs(release): 2.3.0 notes say what is reachable and what is library-only (Refs #177)
v2.3.0
2026-10-11 15:55:12 -04:00
John SmithandClaude Opus 5.5 8f6def6f83 docs(release): 2.3.0 notes say what is reachable and what is library-only (Refs #177)
The 2.0.0-2.3.0 entries described the Phase 6-9 mechanisms as working
flows ("stop needing a human passphrase", "hot-swap its own binaries").
A symbol search over packages/*/src finds no caller for the machine key,
the staging pipeline, the prompt optimizer, the canary runner or lease
acquisition, and run-dag plus the canary live in the private fleet package.

- CHANGELOG: each intro now names what is reachable; 2.3.0 lists reachable,
  library-only and not-in-npm pieces, and records that 1.6.1-2.2.0 were
  never published and ship in 2.3.0.
- README status line: same correction.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-11 15:37:38 -04:00
BillyOutlast c886f66210 Merge pull request #314 from Heretek-AI/fixup/research-test-hermetic
fix(opencode): keep ambient search keys out of the research host tests (Refs #312)
2026-10-11 15:36:10 -04:00
BillyOutlast 1a58c5d655 Merge pull request #313 from Heretek-AI/phase/6-clean-room
fix(tooling): scope test, lint and typecheck to packages/ (Refs #159)
2026-10-11 15:36:05 -04:00
John SmithandClaude Opus 5.5 2dce039f63 fix(opencode): keep ambient search keys out of the research host tests (Refs #312)
With BRAVE_API_KEY exported in the developer shell, fires-research and
research booted hosts whose provider chain reached the live Brave API
before the local SearXNG fake: six tests went red and each run spent the
real key. Both files now remove BRAVE_API_KEY and FIRECRAWL_API_KEY for
their lifetime and restore them afterwards, as failover-chain.test.ts
already does. The degradation block closes through bootHost's handle so
the restore runs there too.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-11 15:18:24 -04:00
John SmithandClaude Opus 5.5 8d6dce56d3 fix(tooling): scope test, lint and typecheck to packages/ (Refs #159)
`bun run check` could not pass in a checkout that holds the gitignored
review/ reference repos. `biome check .` stopped on their nested root
configs, and the bare `packages/web` test filter also matched
review/*/packages/web, so third-party suites ran and the workspace suite
never started. CI and ticket worktrees have no review/, which hid it.

- package.json: the test script passes ./packages/<dir> paths, not filters.
- bunfig.toml: test discovery starts at packages/.
- biome.json: force-ignore review, evals/results and nested checkout roots.
- tsconfig.base.json: exclude review and evals/results.
- core/test/clean-room.test.ts: runs the real root config against a tree
  with planted intruders, one test per layer.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-11 15:18:24 -04:00
BillyOutlast 6c7a62381a Merge pull request #311 from Heretek-AI/phase/9-release-2.3.0
Release 2.3.0: Recursive Self-Improvement & Model Distillation (#177)
2026-10-11 05:15:01 -04:00
John Smith 6db94e44d8 chore(release): 2.3.0 version bump, changelog and status line (Refs #177) 2026-10-11 05:09:03 -04:00
BillyOutlast e8672e8398 Merge pull request #310 from Heretek-AI/phase/9-canary
Phase 9: canary self-compilation and hot-swap harness (#176)
2026-10-11 05:08:54 -04:00
BillyOutlast 444d033488 Merge pull request #309 from Heretek-AI/phase/9-harvester
Phase 9: trajectory harvester for distillation (#175)
2026-10-11 05:08:43 -04:00
John Smith 1629bbbe09 fix(improve): split planted credential name/value (Codacy hardcoded-password, Refs #175) 2026-10-11 05:03:31 -04:00
John Smith 15e83812fc fix(improve): interpolated test secrets, hoisted JSON payloads, .at(-1) (Codacy, Refs #175) 2026-10-11 05:03:31 -04:00
John Smith d498839b16 feat(improve): trajectory harvester with SFT/DPO export-dataset (Refs #175) 2026-10-11 05:03:31 -04:00
BillyOutlast b067f4f4da Merge pull request #308 from Heretek-AI/phase/9-optimizer
Phase 9: evolutionary prompt optimizer (#174)
2026-10-11 05:02:56 -04:00
John Smith a68040ba79 feat(fleet): canary self-compilation and hot-swap harness, off by default (Refs #176) 2026-10-11 04:46:16 -04:00
John Smith 2337e99395 fix(improve): hoist JSON payloads out of template literals (Codacy FP, Refs #174) 2026-10-11 04:29:10 -04:00
John Smith 5202a8ae44 fix(improve): regex-free slug (CodeQL ReDoS) + proposal-writing tests (Refs #174) 2026-10-11 04:26:46 -04:00
John Smith c4bbff8116 feat(improve): evolutionary prompt optimizer with Pareto gate and human-gated apply (Refs #174) 2026-10-11 04:20:27 -04:00
BillyOutlast d95e76c82b Merge pull request #307 from Heretek-AI/phase/8-release-2.2.0
Release 2.2.0: Anti-Goodhart Verification & Mutation Oracle (#172)
2026-10-11 04:15:30 -04:00
John Smith 39ce9e477f style(changelog): blank lines around 2.2.0 lists (Refs #172) 2026-10-11 04:08:37 -04:00
John Smith 3b87431024 chore(release): 2.2.0 version bump, changelog and status line (Refs #172) 2026-10-11 04:08:37 -04:00
BillyOutlast 8c2cce8d32 Merge pull request #306 from Heretek-AI/phase/8-benchmarks
Phase 8: rigorous SWE benchmark suite (#171)
2026-10-11 04:08:27 -04:00
BillyOutlast e03ade0ea8 Merge pull request #305 from Heretek-AI/phase/8-model-policy
Phase 8: heterogeneous multi-model policy (#170)
2026-10-11 04:08:17 -04:00
BillyOutlast 906cfd2063 Merge pull request #304 from Heretek-AI/phase/8-mutation-gate
Phase 8: AST mutation testing gate (#169)
2026-10-11 04:08:06 -04:00
John Smith 2bebbd25ce refactor(evals): split benchmark catalog per category (Refs #171) 2026-10-11 03:41:46 -04:00
John Smith e82dc3177e refactor(evals): split benchmark grader to lower complexity (Refs #171) 2026-10-11 03:38:37 -04:00
John Smith 5ff3b46971 feat(evals): rigorous SWE benchmark suite with oracle grading (Refs #171) 2026-10-11 03:29:42 -04:00
John Smith 802a9368ec test(models): avoid stringify helper in host directives (Refs #170) 2026-10-11 03:13:45 -04:00
John Smith 5a98766e58 fix(policy): tolerate runtimes without config in seat launch guard (Refs #170) 2026-10-11 03:03:52 -04:00
John Smith 727ac35326 refactor(models): split diversity enforcement into agents/models.ts (Refs #170) 2026-10-11 02:47:55 -04:00
John Smith d51110544c feat(models): heterogeneous multi-model policy enforcement (Refs #170) 2026-10-11 02:40:56 -04:00
John Smith 6545be1ca9 refactor(gates): split mutation runners to lower complexity (Refs #169) 2026-10-11 02:30:13 -04:00
John Smith c16c33ca0f style(gates): import order (Refs #169) 2026-10-11 02:28:17 -04:00
John Smith c1e18d6362 fix(gates): constant-time mutation cache compare (Refs #169) 2026-10-11 02:28:08 -04:00
John Smith d901915f86 fix(gates): test-aware mutation cache fingerprint (Refs #169) 2026-10-11 02:20:48 -04:00
John Smith c95091ea27 feat(gates): AST mutation testing gate with kill-rate rejection (Refs #169) 2026-10-11 02:19:38 -04:00
BillyOutlast 543b1d0801 Merge pull request #303 from Heretek-AI/phase/7-release-2.1.0
Release 2.1.0: Autonomous Key Delegation & Consensus Governance (#167)
2026-10-11 02:12:19 -04:00
John Smith b9e41e1e37 chore(release): 2.1.0 version bump, changelog and status line (Refs #167) 2026-10-11 02:02:24 -04:00
BillyOutlast 889b61035a Merge pull request #302 from Heretek-AI/phase/7-staging-pipeline
Phase 7: Autonomous staging pipeline with human release veto (#166)
2026-10-11 02:02:14 -04:00
BillyOutlast cb5ebd0916 Merge pull request #301 from Heretek-AI/phase/7-semantic-rebase
Phase 7: Automated semantic rebase & conflict resolver (#165)
2026-10-11 02:02:04 -04:00
BillyOutlast e213a97f61 Merge pull request #299 from Heretek-AI/phase/7-machine-keystore
Phase 7: MachineKeyStore + proof-carrying transitions (#164)
2026-10-11 02:01:53 -04:00
BillyOutlast ceb9bcc970 Merge pull request #300 from Heretek-AI/fixup/remediation-iv
Remediation IV: fork SessionStart wiring + scout advisory probe enforcement (Refs #289, Refs #257)
2026-10-11 02:01:43 -04:00
John Smith 3350b3dc85 test(factory): dedupe staging setups below gate threshold (Refs #166) 2026-10-11 01:44:18 -04:00
John Smith 31c4de0214 fix(factory): isolated parallel suite checkouts (Refs #166) 2026-10-11 01:36:15 -04:00
John Smith 411fcced0b style(worktree): for-of over indexed loop (Refs #165) 2026-10-11 01:35:47 -04:00
John Smith 45bbb2c3ba fix(worktree): test fast-forwards, fatal continue, default+named imports (Refs #165) 2026-10-11 01:35:24 -04:00
John Smith b805a5fe5c fix(factory): clear Sonar findings on staging pipeline (Refs #166) 2026-10-11 01:34:08 -04:00
John Smith 10da5ca78f fix(worktree): clear remaining Sonar findings, dedupe tests (Refs #165) 2026-10-11 01:33:34 -04:00