Files
gstack/scripts/brain-cache-spec.ts
T
Garry Tan 943105f109 v1.91.8.0 test: smaller suite, derived paid selection, retired never-green evals (#2994)
* test: delete test-infrastructure dead code (G)

- exit-propagation drives the runner's real strict verdict
  (BunTestOutputClassifier + strictTestExitCode); delete the unused
  shardRunLooksTruncated predicate.
- delete skill-coverage-matrix registry + its gate (nothing reads it; the
  floor already iterates skillCensus()).
- delete touchfiles-facade export-parity tests (Bun fails missing imports
  at link time) and the duplicated E2E_TIERS tier-value test.
- delete brain-cache-spec TRANSPORT_DEFAULT_POLICY, SKILL_RUN_RETENTION_DAYS
  and the now-unused BrainTrustPolicy type with their literal tests.
  AUTOPLAN_PREFLIGHT_BUDGET_BYTES stays: skill-preflight-budget enforces it
  against real resolver output.
- delete audit-compliance's JSDoc-comment grep.

* test: replace product tests that fake the product with real-boundary tests (F)

- design: serve.test.ts drove an inline mirror server; now two tests run the
  real serve() on an ephemeral port (reload confinement, submit exit 0).
- setup-gbrain: rollback + voyage tests execute the template-extracted init
  blocks (3 sites) instead of drifted local bash copies.
- terminal-agent: internalHandler source greps replaced by a behavioral
  /internal/grant + /internal/revoke auth matrix (no/wrong/valid token).
- /health: server-security-surface and the server-auth / security-audit-r2 /
  sidebar-tabs source greps fold into one liveness-only check on the real
  body; the L4 sidecar wiring gets a behavioral /pty-inject-scan test.
- delete tautologies (browser-manager onDisconnect, memory-command #12),
  ios swiftui tap fixture self-check, memory-ingest put_page grep, detach
  source greps, sidebar-agent absence pins, dead-CSS pins + the dead CSS,
  security-audit-r2 Task 1 + the test-only meta-commands re-export,
  duplicate generated-SKILL.md checks.
- make-pdf coverage-gaps cases move into their owner test files.

* test: delete tests of dead eval code (A)

- A1: the retired Eng lexical oracle (evaluateEngSeedCoverage,
  isEngSeedDecisionAUQ), the completion-handoff detector and the retained
  corpus had no paid caller since v1.87.6; delete their 26 replay files,
  ~2.6k helper LOC and fixtures, and the dead blocks in 8 mixed files
  (live hasNativePlanTerminal / batching assertions stay).
- A2: dead viewport approvers in autoplan-artifact-permission and their 11
  replay files + fixtures; recorder/launcher cases stay.
- A3: never-wired oracles and seeders (autoplan-phase-order,
  eng-finding-fixture, ceo-paired-fixture, design-ui-scope,
  plan-skill-completion, pty-current-screen, required-reads,
  transcript-section-logger); plan-seed-submission now decodes through the
  production createPtyScreen; section manifests name their actual guard.
- A4: zero-reference helper exports, plus execGit and invokeAndObserve
  found by the reachability pass.
- 52 fixtures orphaned by the deletions; touchfile and selection-table
  entries for every deleted path.

* test: clean up the paid eval lane (B1-B4, B6, B7)

- B1: delete paid files that assert nothing or cannot pass meaningfully:
  skill-llm-eval-spec and skill-e2e-spec-execute (test.todo), gemini-e2e
  (+ gemini-session-runner; no gemini CLI in CI), ship-idempotency (red
  since v1.63), the two opus-4-7 *-sonnet overlay wrappers, conductor-prose
  (+ its source-evaluation replay), codex-e2e-plan-format; drop their keys,
  scripts and census rows.
- B2: skill-llm-eval grades browse/sections/command-list.md with one union
  judge that also carries the baseline score pin; regression-vs-baseline
  deleted (paid run: pass, c4/c4/a4).
- B3: memory-pipeline, ios-qa, ios-qa-swift-build and plan-tune-cathedral
  make no model calls; renamed out of the paid glob so they run on every
  PR. Swift builds need GSTACK_TEST_SWIFT=1; device stub deleted.
- B4: codex-e2e*, outside-voice, aside and ios-device cannot run in the CI
  image; excluded from the weekly lane with a tracked re-entry condition.
- B6: fold opus-47's negative routing controls into skill-routing-e2e
  journey-negatives (paid run: 3/3 unrouted) and delete the file.
- B7: delete the never-green brain-privacy-gate eval; a free
  gstack-skill-start test now proves consent precedes artifacts egress.

* test: retire the finding-count cluster and trim its helpers (C)

- C0/C1: the five never-green evals (skill-e2e-autoplan-chain and
  skill-e2e-plan-{ceo,eng,design,devex}-finding-count) failed on harness and
  budget, never on skill behavior; delete them, their touchfile/tier ids,
  AUTOPLAN_CHAIN_BUDGET and the dedicated eighth periodic slice (--slices 7).
- C2: delete the helper groups whose only paid consumers were those files
  (11 modules), trim claude-pty-runner and eng-seeded-coverage to the paid
  closure, and delete the free replay tests whose assertions exercised only
  that dead code (89 files, 135 orphaned fixtures). Blocks that used dead code
  only as input for a live subject keep their assertions: the multiSelect
  default moved to plan-review-decisions, runner PTY tests use inline caller
  policies, and the timer-safe budget checks moved to eng-finding-retry-budget.
- The eight production-touching files stay except ceo-current-decision-record
  (its template read only feeds the retired counter).
- CARVE_GUARDS.autoplan is behavioral 'none'; TODOS records the lost chain
  and per-finding cadence coverage with their re-entry tests.

* test: fold per-incident replay series into their detector owners (D)

Twelve detector families move into one owner test each: 73 incident files
become describe blocks in ceo-section-loading-fixture (stale-fill race),
model-overlays, coverage-audit-evidence, autoplan-phase-observer,
native-auto-decide, outside-voice-evidence, eng-first-review,
plan-count-completion, plan-count-file-permission, ceo-mode-option,
plan-scope-selection and plan-count-prerequisite. Each block keeps its original
code and fixture, so every case still runs; only tests asserting the incident
file's own touchfile registration are dropped (41). Touchfile lists that named
an incident now name its owner.

* test: start the plan-count history PTY on its readiness marker (H)

The fake CLI prints a startup marker and the runner waits for it instead of the
fixed 8 s startup sleep (8.6 s -> 0.9 s locally). eng-semantic-terminal's
sleeping registration cases went with C; plan-count-timeout keeps the fixed wait
because it asserts deadline behavior.

* test: derive paid touchfiles from each eval's static closure (E)

touchfiles.test.ts now checks, per key, that the paid file's static
test/helpers and test/fixtures closure (plus fixture paths it names in string
literals) is covered, and names the file, path, chain and key to fix when it is
not. Free *.test.ts files are no longer touchfiles, so editing a free replay
test stops selecting paid evals: 950 entries removed, 653 real closure paths
added. The hand-copied inventories go: periodic-fixture-selection,
fake-impeccable-touchfiles and 45 per-file selection examples. Selection for
the sample edits (plan-eng-review template, claude-pty-runner,
plan-count-fixture, gstack-config) loses no case under either profile.
CONTRIBUTING documents the rule and its lower bound.

* test: skip hollow tier shards and census judges in the paid planner (B5)

A paid file is now skipped for a tier lane only when every E2E id it registers
is known statically and none has that tier; ids come from the touchfile
registrations and literal testName/*IfSelected arguments, so a comment or
skill path that quotes another id cannot unschedule it, and computed names
keep today's scheduling. --list and the manifest show each skip as
"skipped: no E2E_TIERS id has tier <tier>". The weekly gate census drops the
LLM judges (--skip-judges); they still run in the periodic census and PR gate
lanes. Gate lane 52 -> 42 files, census 41; periodic 77 -> 69.

* test: run seven paid evals on the current default capture model (B8)

skill-e2e-{auq-matrix,plan-format,qa-bugs,retro,workflow} pinned
claude-opus-4-7 and skill-e2e-office-hours plus -brain-writeback pinned
claude-sonnet-4-6; none tests a historical model, so they now capture with
resolveEvalModel('capture'), and the free harness tests that execute these
registrations receive the same resolver. The paid re-pin run passed all of
them. skill-e2e-{design,office-hours-phase4,plan-prosons,plan} keep
claude-opus-4-7: six of their cases failed on the default model (three
timeouts, a missing report file, a format miss and a posture score of 3), so
per the plan's fallback they keep their pins with a TODOS entry. The pre-spend
estimate and drop threshold are in docs/test-audit-2026-09.md.

* test: guard the reduced suite against new test-of-test files

- test/test-of-test-ratchet.test.ts records the 228 free tests that import only
  test/ code and fails on a new one, naming the owner test to extend instead;
  a stale baseline entry fails with the remove instruction.
- test/helpers/resolve-repo-path.ts is the one specifier/literal resolver for
  the ratchet and the touchfile closure invariant, with its own unit tests.
- CONTRIBUTING "Test tiers" describes the paid-failure workflow (fix, then one
  row in the detector's owner test) and the ratchet; TEST_PORTFOLIO gains the
  detector -> owner-test table and no longer claims an Autoplan chain eval.
- TODOS: automatic exclusion policy for chronically red periodic files (P3),
  the deferred native-completion table collapse, the unused CEO payment
  seeder; the PTY readiness item is narrowed to the paid runner.
- docs/test-audit-2026-09.md collects the triage, security mapping, inventories,
  selection proof, behavior-commit decisions and retained false positives.

* v1.91.8.0 test: smaller suite, derived paid selection, retired never-green evals

Release metadata for the test-reduction branch: VERSION 1.91.8.0 (1.91.7.0 is
claimed by #2983), CHANGELOG with the measured before/after table and a
contributor section, durations re-recorded on Ubicloud standard-16 (857 files,
0 failures), the agents digest, CONTRIBUTING's after-measurement row, the B8
fallback TODOS entry, and the after metrics, kept-vs-plan notes, B8 run and
census estimate in docs/test-audit-2026-09.md.

* fix(ubicloud): skip retrieval globs that match nothing instead of reporting a failed pull

* test: count issue-numbered Design findings in the UI-scope gate eval
2026-09-29 08:58:49 -07:00

243 lines
9.0 KiB
TypeScript

/**
* Brain cache spec — single source of truth for the brain-aware planning skills
* cache layer. Imported by:
* - scripts/resolvers/gbrain.ts (renders per-skill subset into SKILL.md.tmpl)
* - bin/gstack-brain-cache (drives TTL + write-back invalidation)
* - test/brain-cache-spec.test.ts (asserts internal consistency)
* - test/skill-preflight-budget.test.ts (enforces per-skill token budget)
* - test/autoplan-preflight-budget.test.ts (enforces autoplan total budget)
*
* Drift between docs and runtime is impossible by construction: the same
* const drives both the rendered table in SKILL.md and the cache CLI behavior.
*/
export interface BrainCacheEntity {
/** Filename inside ~/.gstack/{,projects/<slug>/}brain-cache/ */
file: string;
/** Time-to-live in milliseconds before cache is considered stale and triggers cold refresh. */
ttl_ms: number;
/** Scope determines which dir holds the cache file. */
scope: 'cross-project' | 'per-project';
/**
* Which write-paths invalidate this digest. When a writer runs, it consults
* this list to know which cache files to bust. Special values:
* - 'calibration-write' — any Phase 2 takes_add call
* - 'skill-run-write' — any skill that writes a gstack/skill-run page
* Otherwise these are skill names like '/plan-ceo-review'.
*/
invalidated_by: ReadonlyArray<string>;
/** Hard byte budget for the digest. Compressor drops oldest items if exceeded. */
budget_bytes: number;
}
/**
* The seven cached entities mirror the seven typed page kinds in
* `gstack-core` schema pack v1.0.0 (Phase 0):
* user-profile, product, goal, developer-persona, brand, competitive-intel, skill-run
* Plus two derived digests:
* recent-decisions (top 5 gstack/skill-run pages)
* salience (mcp__gbrain__get_recent_salience output)
*/
export const BRAIN_CACHE_ENTITIES: Record<string, BrainCacheEntity> = {
'user-profile': {
file: 'user-profile.md',
ttl_ms: 7 * 86_400_000, // 7 days
scope: 'cross-project',
invalidated_by: ['/retro', '/plan-tune', 'calibration-write'],
budget_bytes: 2048,
},
product: {
file: 'product.md',
ttl_ms: 1 * 86_400_000, // 1 day
scope: 'per-project',
invalidated_by: ['/office-hours', '/plan-ceo-review'],
budget_bytes: 1024,
},
goals: {
file: 'goals.md',
ttl_ms: 12 * 3_600_000, // 12 hours
scope: 'per-project',
invalidated_by: ['/office-hours', '/plan-ceo-review'],
budget_bytes: 512,
},
'developer-persona': {
file: 'developer-persona.md',
ttl_ms: 7 * 86_400_000,
scope: 'per-project',
invalidated_by: ['/plan-devex-review', '/devex-review'],
budget_bytes: 1024,
},
brand: {
file: 'brand.md',
ttl_ms: 7 * 86_400_000,
scope: 'per-project',
invalidated_by: ['/design-consultation', '/plan-design-review'],
budget_bytes: 1024,
},
'competitive-intel': {
file: 'competitive-intel.md',
ttl_ms: 1 * 86_400_000,
scope: 'per-project',
invalidated_by: ['/plan-ceo-review', '/office-hours'],
budget_bytes: 1024,
},
'recent-decisions': {
file: 'recent-decisions.md',
ttl_ms: 12 * 3_600_000,
scope: 'per-project',
invalidated_by: ['skill-run-write'],
budget_bytes: 2048,
},
salience: {
file: 'salience.md',
ttl_ms: 4 * 3_600_000, // 4 hours
scope: 'per-project',
invalidated_by: [],
budget_bytes: 512,
},
};
/**
* Per-skill subset map. The resolver consumes this to emit per-skill BRAIN_PREFLIGHT
* instructions. The skill template loads ONLY the listed digests — never more.
* Order matters for narrative coherence in the injected ## Brain Context block.
*
* Hard token budget per skill (validated by test/skill-preflight-budget.test.ts):
* - CEO/office-hours: 5 KB (richest context need)
* - eng/design/devex: 2 KB
*/
export const SKILL_DIGEST_SUBSETS: Record<string, ReadonlyArray<string>> = {
'office-hours': ['product', 'goals', 'user-profile', 'recent-decisions', 'salience'],
'plan-ceo-review': ['product', 'goals', 'recent-decisions', 'user-profile'],
'plan-eng-review': ['product', 'recent-decisions'],
'plan-design-review': ['product', 'brand', 'recent-decisions'],
'plan-devex-review': ['product', 'developer-persona', 'recent-decisions', 'competitive-intel'],
};
/** Per-skill total digest budget (sum of loaded digests must not exceed). */
export const SKILL_PREFLIGHT_BUDGET_BYTES: Record<string, number> = {
'office-hours': 5120,
'plan-ceo-review': 5120,
'plan-eng-review': 2048,
'plan-design-review': 2048,
'plan-devex-review': 2048,
};
/**
* Total budget across an autoplan run (4 sequential planning skills). Validated by
* test/autoplan-preflight-budget.test.ts. If a future autoplan-extended adds skills,
* this cap forces an explicit budget revisit.
*/
export const AUTOPLAN_PREFLIGHT_BUDGET_BYTES = 25_600;
/**
* D9 salience privacy: default allowlist of slug prefixes that are safe to surface
* in planning prompts. Anything outside (personal/, family/, therapy/, etc.)
* gets stripped at digest write time. User can extend via
* `gstack-config set salience_allowlist '<comma-separated-prefixes>'`.
*/
export const SALIENCE_DEFAULT_ALLOWLIST: ReadonlyArray<string> = [
'projects/',
'concepts/',
'gstack/',
];
/**
* Per-skill calibration bet weights (Phase 2 / E5). When a planning skill writes
* a kind=bet take, the weight determines how strongly it factors into the user's
* calibration profile. Higher = more confident prediction worth more credit/blame
* on resolution.
*/
export const SKILL_CALIBRATION_WEIGHTS: Record<string, number> = {
'plan-ceo-review': 0.8,
'plan-eng-review': 0.7,
'plan-design-review': 0.5,
'plan-devex-review': 0.6,
'office-hours': 0.9,
};
/**
* Lock-file path used by the cache refresh dedup (D3). Per-project to avoid
* cross-project contention. Stale-takeover after 5 minutes.
*/
export const CACHE_REFRESH_LOCK_TIMEOUT_MS = 5 * 60_000;
/**
* Schema pack identity. Bumped when adding/removing/renaming page types.
* On mismatch with the version recorded in _meta.json, the cache layer
* triggers a FULL rebuild for the affected project.
*/
export const GSTACK_SCHEMA_PACK_NAME = 'gstack-core';
export const GSTACK_SCHEMA_PACK_VERSION = '1.0.0';
/**
* User-slug fallback chain (D4 A3 defensive default). Resolved once per endpoint
* and persisted via `gstack-config set user_slug_at_<endpoint-hash> <slug>`.
* Stable across sessions.
*/
export const USER_SLUG_RESOLUTION_ORDER = [
'whoami_client_name', // mcp__gbrain__whoami.client_name (remote + OAuth)
'env_user', // $USER environment variable
'git_email_sha8', // sha8($(git config user.email))
'anonymous_hostname_sha8', // anonymous-<sha8(hostname)>
] as const;
/** ----------------------------------------------------------------------- */
/** Helper functions consumed by the resolver, cache CLI, and tests. */
/** ----------------------------------------------------------------------- */
/** Returns the cache filename for an entity name, throws if unknown. */
export function getCacheFile(entityName: string): string {
const entity = BRAIN_CACHE_ENTITIES[entityName];
if (!entity) throw new Error(`Unknown brain cache entity: ${entityName}`);
return entity.file;
}
/** Returns the digest subset for a skill, throws if the skill isn't preflight-enabled. */
export function getSkillSubset(skillName: string): ReadonlyArray<string> {
const subset = SKILL_DIGEST_SUBSETS[skillName];
if (!subset) throw new Error(`Skill not registered for brain preflight: ${skillName}`);
return subset;
}
/** Returns the per-skill total digest budget in bytes. */
export function getSkillBudget(skillName: string): number {
const budget = SKILL_PREFLIGHT_BUDGET_BYTES[skillName];
if (budget == null) throw new Error(`Skill not registered for brain preflight: ${skillName}`);
return budget;
}
/**
* Given a write-path identifier (skill name or special token), returns the list
* of cache files that should be invalidated. Drives the cache CLI's `invalidate`
* subcommand and the resolver's BRAIN_WRITE_BACK block.
*/
export function getInvalidationTargets(writePath: string): ReadonlyArray<string> {
const targets: string[] = [];
for (const [name, entity] of Object.entries(BRAIN_CACHE_ENTITIES)) {
if (entity.invalidated_by.includes(writePath)) {
targets.push(name);
}
}
return targets;
}
/**
* Lists all skill names that are registered for brain preflight. Used by
* test/brain-preflight.test.ts and test/skill-preflight-budget.test.ts to
* iterate without hardcoding the skill list.
*/
export function getPreflightSkills(): ReadonlyArray<string> {
return Object.keys(SKILL_DIGEST_SUBSETS);
}
/**
* Computes the maximum possible digest set size for a skill (sum of per-entity
* budgets in the subset). Used by skill-preflight-budget.test.ts to validate
* that the per-skill cap is enforceable given the per-entity caps.
*/
export function getMaxSubsetBytes(skillName: string): number {
const subset = getSkillSubset(skillName);
return subset.reduce((sum, name) => sum + (BRAIN_CACHE_ENTITIES[name]?.budget_bytes ?? 0), 0);
}