Files
zero/internal/providers/compaction_model_test.go
Kevin CodexandClaude Fable 5 6645f9744b perf(agent): cheap compaction — dedicated summarizer model, input cap… (#532)
* perf(agent): cheap compaction — dedicated summarizer model, input caps, free reactive prune, window re-derive

Compaction summarization was the single most expensive recurring event
in a long run: the full elided middle — verbatim tool bodies included —
went to the session's main model at full price (a ~100k-token call near
a 200k window, recursively multiplied on overflow).

Four changes, each independently safe:

- Dedicated summarizer model. Compaction summaries route to a cheap
  model resolved as ZERO_COMPACTION_MODEL > preferences.compactionModel
  > a curated default on official endpoints (Haiku 4.5 / Gemini
  Flash-Lite / gpt-4.1-mini; custom-endpoint catalogs are unknowable so
  they get none). Built lazily on the first paid compaction; ANY
  failure permanently falls back to the main provider for the run, so a
  misconfigured summarizer can never break compaction. "main" at any
  level forces the old behavior.

- Summarizer input caps. renderTranscript clamps tool-result bodies to
  2,000 chars and tool-call arguments to 500 (head+tail, rune-safe) —
  the summary needs what was done and decided, not raw payloads. Cuts
  summarizer input 5-10x on tool-heavy sessions.

- Free reactive prune. recover() now runs the same prune-first stage as
  the proactive path: a context-limit error is first answered by
  pruning stale tool bodies at zero cost, without consuming the
  one-shot reactive budget; the paid summarizer only runs when a later
  error finds nothing left to prune.

- Window re-derivation on model switch. Options.ContextWindowFor
  (wired to the model registry in exec and the TUI) lets the loop
  re-derive the compaction threshold after escalate_model, closing the
  documented KNOWN LIMITATION where a switch to a smaller-window model
  could overflow and a larger one over-compacted.

## Rebase notes (2026-08-25, onto 2749447c)

Rebased across 214 commits. main now summarizes through the context planner
(#879) and switches models via turn sessions, so the agent side was reworked
rather than merged textually:

- Summarizer selection lives on compactionState and is resolved lazily per
  main model through summarizeWithPlanner (planner + trace preserved); a
  failing dedicated summarizer is marked broken and the same slice is
  retried once on the main provider.
- Options.Summarizer now receives the main model in force:
  `func(ctx, mainModelID string) (Provider, error)`, and may return
  (nil, nil) when no dedicated model applies. After a successful
  escalate_model switch the loop calls compactor.switchModel, which
  re-derives the window and re-resolves the summarizer — a run that started
  on the cheap model and escalated regains cheap summaries (review P2).
- recover()'s free prune stage runs before the one-shot reactive guard
  (pruning is free and cannot loop) and mirrors maybeCompact's two prune
  passes; the free retry is taken only when the pruned history fits the
  window the provider enforces (with a 5% estimate margin), otherwise the
  same call falls through to the paid summarizer. Gating on the 70%
  proactive threshold would have sent histories in the 70-95% band to the
  summarizer unnecessarily (review P1 and the follow-up question).
- ContextWindowFor is wired through the turn-session switch path; the
  planner's window is updated alongside the compactor's.

Tests: the PR's suite is ported to the new seams, plus insufficient-prune
fall-through, free prune after the budget is spent, window-vs-threshold
gating, and summarizer re-resolution across a model switch.

* fix(agent): branch on a dedicated-summarizer flag instead of comparing providers

Review follow-ups for #532:

- summarizeProvider now reports whether the dedicated summarizer was
  selected and summarizeClosure branches on that flag; the previous
  `chosen == provider` compared Provider interface values, which panics
  when the dynamic type is not comparable. Regression test uses a provider
  with a slice field for both the summarizer and the main provider.
- CompactionModelID's curated-default test covers the official OpenAI
  profile alongside Anthropic and Google.

Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_012JqvAf7zsP1qYFVZayKXHq

---------

Co-authored-by: Claude Fable 5 <[email protected]>
2026-08-25 21:18:04 +08:00

114 lines
4.7 KiB
Go

package providers
import (
"context"
"errors"
"testing"
"github.com/Gitlawb/zero/internal/config"
"github.com/Gitlawb/zero/internal/zeroruntime"
)
func TestCompactionModelIDResolutionOrder(t *testing.T) {
anthropic := config.ProviderProfile{
Name: "anthropic",
ProviderKind: config.ProviderKindAnthropic,
Model: "claude-sonnet-4.5",
}
t.Setenv(CompactionModelEnv, "my-env-model")
if got := CompactionModelID(anthropic, "my-config-model"); got != "my-env-model" {
t.Fatalf("env must win, got %q", got)
}
t.Setenv(CompactionModelEnv, "main")
if got := CompactionModelID(anthropic, "my-config-model"); got != "" {
t.Fatalf("env 'main' must force the main model, got %q", got)
}
t.Setenv(CompactionModelEnv, "")
if got := CompactionModelID(anthropic, "my-config-model"); got != "my-config-model" {
t.Fatalf("config preference must apply when env unset, got %q", got)
}
if got := CompactionModelID(anthropic, "main"); got != "" {
t.Fatalf("config 'main' must force the main model, got %q", got)
}
}
func TestCompactionModelIDCuratedDefaults(t *testing.T) {
t.Setenv(CompactionModelEnv, "")
anthropic := config.ProviderProfile{Name: "anthropic", ProviderKind: config.ProviderKindAnthropic, Model: "claude-sonnet-4.5"}
if got := CompactionModelID(anthropic, ""); got != defaultAnthropicCompactionModel {
t.Fatalf("anthropic default = %q, want %q", got, defaultAnthropicCompactionModel)
}
google := config.ProviderProfile{Name: "google", ProviderKind: config.ProviderKindGoogle, Model: "gemini-2.5-pro"}
if got := CompactionModelID(google, ""); got != defaultGoogleCompactionModel {
t.Fatalf("google default = %q, want %q", got, defaultGoogleCompactionModel)
}
openai := config.ProviderProfile{Name: "openai", ProviderKind: config.ProviderKindOpenAI, Model: "gpt-4.1"}
if got := CompactionModelID(openai, ""); got != defaultOpenAICompactionModel {
t.Fatalf("openai default = %q, want %q", got, defaultOpenAICompactionModel)
}
// Custom/compatible endpoints: catalog unknowable, no default.
compatible := config.ProviderProfile{
Name: "opengateway",
ProviderKind: config.ProviderKindOpenAICompatible,
BaseURL: "https://opengateway.gitlawb.com/v1",
Model: "some-model",
}
if got := CompactionModelID(compatible, ""); got != "" {
t.Fatalf("openai-compatible must have no default, got %q", got)
}
}
func TestCompactionModelIDSkipsWhenMainIsAlreadyCheap(t *testing.T) {
t.Setenv(CompactionModelEnv, "")
haiku := config.ProviderProfile{
Name: "anthropic",
ProviderKind: config.ProviderKindAnthropic,
Model: defaultAnthropicCompactionModel,
}
if got := CompactionModelID(haiku, ""); got != "" {
t.Fatalf("already-cheap session must not get a dedicated summarizer, got %q", got)
}
}
func TestCompactionSummarizerFactoryTracksMainModel(t *testing.T) {
t.Setenv(CompactionModelEnv, "")
base := config.ProviderProfile{Name: "anthropic", ProviderKind: config.ProviderKindAnthropic, Model: defaultAnthropicCompactionModel}
var built []string
newProvider := func(profile config.ProviderProfile) (zeroruntime.Provider, error) {
built = append(built, profile.Model)
return nil, errors.New("not needed")
}
factory := CompactionSummarizerFactory(base, "", newProvider)
if factory == nil {
t.Fatal("factory must exist when a provider builder exists")
}
// Main model is the cheap default: no dedicated summarizer, nothing built.
if provider, err := factory(context.Background(), defaultAnthropicCompactionModel); provider != nil || err != nil || len(built) != 0 {
t.Fatalf("cheap main model must yield (nil, nil) without building, got %v %v built=%v", provider, err, built)
}
// After escalating to an expensive model the cheap default is built.
if _, err := factory(context.Background(), "claude-sonnet-4-5"); err == nil || len(built) != 1 || built[0] != defaultAnthropicCompactionModel {
t.Fatalf("escalated main model must build the cheap summarizer, built=%v err=%v", built, err)
}
// An empty main model falls back to the base profile's model — the cheap
// default here, so again nothing is built.
built = nil
if provider, err := factory(context.Background(), ""); provider != nil || err != nil || len(built) != 0 {
t.Fatalf("empty main model must use the base profile, got %v %v built=%v", provider, err, built)
}
// The explicit "main" preference disables the dedicated summarizer.
if provider, err := CompactionSummarizerFactory(base, "main", newProvider)(context.Background(), "claude-sonnet-4-5"); provider != nil || err != nil {
t.Fatalf("preference main must disable the summarizer, got %v %v", provider, err)
}
if CompactionSummarizerFactory(base, "", nil) != nil {
t.Fatal("no provider builder means no factory")
}
}