6 Commits

Author	SHA1	Message	Date
Tom Boucher	596540f864	feat(#3227): publish machine-readable state contract at step boundaries (#3824) ... * feat(#3227): publish machine-readable state contract at step boundaries Adds src/state-contract.cts, a best-effort publisher that writes .planning/state.json (contract 1.0.0) at 11 step-boundary commands, so external tools read a versioned contract instead of parsing STATE.md and ROADMAP.md heuristically. Composes existing owners rather than re-deriving: phase rows come from a new locateProgressTable extracted from deriveProgressFromRoadmap (so the snapshot can never disagree with GSD's own progress counters), milestone identity from getMilestoneInfo, and next from classifyProject. Owners are required lazily to avoid the state -> state-contract -> smart-entry -> state require cycle. Also fixes a pre-existing defect in scripts/lint-test-file-count.cjs (maintainer-approved as a second concern): testEffectivePrefix never stripped the suite qualifier, so 65 dotted test files counted against no module and 9 mis-bucketed into a shorter one. Allowlist re-baselined for the 74 files the gate can now see. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(#3227): backfill PR number into the changeset fragment pr:0 -> pr:3824 now that the PR exists. Doc-only. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test(#3227): shape hostile-name fixtures away from the scan corpus The two hostile-input fixtures used a literal phrase from scripts/prompt-injection-scan.sh's corpus, so CI's Security Scan redded on this file. These tests assert that an arbitrary phase name round-trips into state.json as inert data -- the property holds for any string, so the injection flavor is illustrative, not load-bearing. Reshaped to a hyphenated fake instruction tag, which stays hostile-looking while matching none of the scanner's patterns. Allowlisting the file was rejected: that mechanism is for suites whose subject IS injection defense, and it would blind the scanner to this whole file permanently. See DEFECT.PROMPT-INJECTION-SCAN-COLLISION. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(#3227): ratchet the state-contract mutation floor to its measured score The module was registered at minScore 50, the ratchet's minimum permitted floor for a newly-registered module whose score had not been measured. This PR's own Stryker shard measured 66.25% (run 32769289750, job 97565813640), so the floor moves to floor(measured) - 1 = 65, per the rule the registry documents. 66.25 is below TARGET_MUTATION_SCORE (80), so this stays a ratchet candidate: raise as the tests improve, never lower. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: sim <sim@local> Co-authored-by: Claude Opus 5 <noreply@anthropic.com>	2026-08-24 17:56:02 -04:00
Tom Boucher	004e9dd741	fix(#3007): resolve Codex reasoning effort per model and make every clamp visible (#3765) ... * test(#3007): failing-first suite for per-model Codex effort capability RED by construction. Binds to behavior renderEffortForRuntime does not yet have: an optional third `model` argument, a per-model advertised-level table, `max` passing through instead of clamping to `xhigh`, `minimal` clamping to `low`, `ultra` rejected outright, and clamp visibility (`requested`/`clamped`/ `reason`) so a downgrade is legible from resolver output rather than silent. Two of these pin defects that exist on next today: - `max` is discarded. Both Codex models whose catalog entries are retrievable (sol, luna) advertise `max`; GSD clamps it to `xhigh` and reports nothing. - `minimal` is emitted to a model that refuses it. providerPresets.openai. haiku.low pairs gpt-5.6-luna with reasoning_effort "minimal", and luna's advertised floor is `low`. GSD is sending a value into a document Codex itself validates. The parity test is what pins that fixed, and it names the offending path/model/effort when it trips. Also corrects tests/model-resolver.test.cjs:351, which asserted renderEffortForRuntime('codex','max').value === 'xhigh' -- the defect pinned as though it were a contract. ADR-443 recorded "Codex has no max" as fact and it was true when written; Codex has since added both `max` and `ultra`. That is a stale premise, so the assertion is corrected here rather than worked around. The property test asserts the invariant the whole change exists for: a rendered effort is always a level the target model actually advertises, or an explicit rejection. There is no third outcome. * fix(#3007): resolve Codex effort per model, and make every clamp visible Codex declares supported_reasoning_levels per MODEL and validates against it, so a single per-runtime capability set cannot be right for all of them. GSD's was wrong in both directions at once. `max` reaches Codex now. ADR-443 recorded "Codex has no max" as fact and clamped max -> xhigh on that basis; it was accurate when written, and Codex has since added both `max` and `ultra`. Every Codex model whose catalog entry is retrievable advertises `max`, so the clamp was discarding a level the provider supports, silently, on the most-used path. `minimal` stops reaching Codex. No Codex model advertises it -- both retrievable entries floor at `low` -- yet providerPresets.openai.haiku.low paired gpt-5.6-luna with reasoning_effort "minimal". GSD was writing a value the receiver validates and refuses into a file the receiver reads. Being unconservative in what you send is the half of Postel's rule with no defensible reading, so that preset is corrected and a parity test pins it. `ultra` is refused rather than laddered. Codex's own catalog calls it "Maximum reasoning with automatic task delegation": at ultra, effective_multi_agent_mode returns Proactive and Codex spawns sub-agents on its own initiative, underneath GSD's orchestration rather than inside it (#2167). It is a mode switch, not a reasoning depth, so it is not added to the universal ladder -- which stays provider-agnostic by ADR-443's design -- and it is rejected even for gpt-5.6-sol, which does advertise it. Clamping it down to `max` was considered and rejected: that silently discards what the user actually asked for. Clamping is now visible. RenderedEffort carries requested/clamped/reason and resolve-execution surfaces them. The previous table clamped correctly but invisibly, so a user asking for `max` on Codex had no way to find out they were getting `xhigh` -- exactly the failure mode the robustness principle's modern critique warns about, and why "be liberal" has to mean "liberal and loud". Also closes a latent trap found while reviewing the implementation: the clamp-up loop walks the ladder upward, and for a future model advertising `ultra` but not `max` it would have selected `ultra` as the clamp target -- re-entering by the back door the mode the rejection above exists to keep out. A clamp may never produce a value that a direct request for that value would refuse. Unreachable with today's catalog, which is why no test caught it; a test now asserts the invariant directly. Signature stability is preserved: the third `model` argument is optional and the two-argument form still resolves, against the family baseline. That form's BEHAVIOR does change for `max` and `minimal`, and it must -- keeping the old answer would have fixed the defect only where a model happened to be threaded through and left it live everywhere else. tests/model-resolver.test.cjs:351 asserted the defect as if it were a contract and is corrected here rather than worked around. * fix(#3007): close every review finding on the Codex effort alignment Two isolated reviewers, correctness and security. Both found the same two blockers, and the per-model work was inert on every surface that matters until this commit. BLOCKER — resolve-execution never passed the model and discarded the clamp. cmdResolveExecution called the two-argument form and emitted only effort_rendered/effort_param/effort_propagation, so the per-model table was unreachable from production code (tests were its only caller) and requested/ clamped/reason were computed and thrown away. Requested outcome 3 names "the effective rendered effort in resolver output" specifically, so the feature was unmet on the exact surface the issue asks for. Now passes the resolved model and emits effort_requested / effort_clamped / effort_clamp_reason, flat, matching the existing key convention rather than introducing a nested object. BLOCKER — the docs described output that did not exist. CONFIGURATION.md showed a nested {"effort": ...} sample; the real result is flat and those keys were absent entirely. A reference doc asserting a JSON path a reader can copy is worse than no doc. Corrected against the actual emitted key set. MAJOR — the argv channel still shipped both original defects. EFFORT_ARGV.codex kept minimal in its supported set and still clamped max down to xhigh, so the invocation-time and install-time channels disagreed about the same runtime's capability: --host codex with max emitted xhigh while the generated TOML said max. This is the repo's documented generative-fix-divergence class, so both tables now cross-reference each other and a parity test fails if they ever diverge again. MAJOR — malformed catalog data failed OPEN and could crash the CLI. A null _baseline became an EMPTY Set that is nonetheless truthy, so the nullish fallback never fired and every effort rendered as null. And a non-array value made the Set constructor throw at module load — model-catalog.cjs is required across the whole CLI, so one bad JSON value killed every command, not just codex effort. Guarded on size and filtered to array values; both degrade to the hardcoded baseline. MAJOR — value widened to a nullable string with two consumers left behind. runtime-artifact-conversion passed it straight into injectEffortFrontmatter (a null effort key in generated frontmatter); install-effort-resolver still declared a non-nullable return, a structural lie that silently defeated null checking. Both corrected, both omitting the key on null — the same posture as 'inherit', where omission means "follow the host default". MAJOR — the per-model table is inert today, and the docs now say so. All three shipped models advertise the same usable range and ultra (sol's only differentiator) is rejected for every model, so no observable output differs by model. The table stays because Codex declares capability per model and the sets are free to diverge — a single per-runtime assumption is precisely what went stale and produced this issue — but overselling it as a visible per-model feature would have been the same class of error as the doc blocker above. Tests: three passed under a full revert and are strengthened rather than deleted, since each guards a real contract (#3533's inherit rule, the undeclared-host rule, off-ladder handling) — they now also assert the clamp-visibility fields, which only exist after this change. The fast-check property is kept for its shrinking, and a deterministic nested loop over the full cross-product now sits beside it so coverage is exhaustive rather than sampled. Also folded in earlier: bin/install.js generated the Codex TOML with the two-arg form and would have written a literal null reasoning effort on the ultra path; CONTEXT.md's Model Catalog Module glossary entry now records CODEX_MODEL_EFFORT. The installer defect was found by the co-change gate, not by a reviewer — install.js is a historical co-change partner of model-catalog.cts that this diff had not touched. * test(#3007): correct assertions that pinned Codex's stale effort premise Thirteen pre-existing tests encoded "Codex has no max" as fact and failed on the shipped commit. Every one is a stale pin, not a defect: each was probed against the built module before its expectation was changed, and none failed for a reason other than this premise correction. Kept as its own commit per CONTRIBUTING — a test-fixture correction made stale by a production change must not ride inside another commit, because the release-sdk hotfix cherry-pick filter routes by subject prefix and a correction buried under the wrong prefix ships a half-state (v1.42.3, #3621). The most valuable one was tests/model-resolver.test.cjs's cross-provider validity invariant, which hardcoded the Codex enum as `minimal|low|medium|high|xhigh` and failed with "real API would 400". That message is now false in both directions: Codex accepts `max`, and rejects `minimal`, which no model advertises. The enum is corrected to `low|medium|high|xhigh|max` and the guard is kept intact — it is exactly the "would the real API refuse this" check worth having, and it was right to fail here. It simply carried the stale fact in its own fixture. Test NAMES were corrected alongside their assertions wherever the name asserted the old behavior — "max is Anthropic-only", "max clamps to xhigh", "minimal passthrough". A renamed test that still claims the old thing is worse than a failing one, and a green test whose name states a falsehood is how the next reader inherits the wrong premise. Both channels are covered: install-time (renderEffortForRuntime, and the generated .toml in install-runtime-artifacts) and invocation-time argv (effort-surface-axis). They were deliberately brought into agreement in this change, so their assertions had to move together. Each site carries a #3007 comment recording that Codex gained max/ultra and that capability is declared per model, so a future reader can tell this was a deliberate premise correction rather than a test bent to fit an implementation. * test(#3007): separate the effort-precedence case from the clamp case The previous stale-assertion pass over-corrected one test. It saw `effort: { default: 'max' }` on codex expecting `effort_rendered: 'xhigh'`, assumed the xhigh came from the max→xhigh clamp #3007 removes, renamed it to "max passes through" and changed the expectation to `max`. The remote runner disagreed. Reproduced against the real CLI: with that config and `gsd-planner`, the resolver emits `effort: "xhigh"`, `effort_requested: "xhigh"`, `effort_clamped: false`. The xhigh is produced by effort-resolution PRECEDENCE — gsd-planner is heavy/opus tier and its routing-tier default outranks `effort.default` — so `max` never reaches the renderer at all. The test says nothing about clamping and never did; it only looked like a clamp pin because both mechanisms happened to yield the same string. Restored to `xhigh` and renamed to say what it actually tests. It now also asserts `effort_clamped === false` and `effort_requested === 'xhigh'`, which is what makes it impossible to mistake for a clamp pin again: those two fields prove the value is what the resolver produced rather than something the renderer downgraded. Before #3007 there was no way to tell the two apart from the output — which is precisely why the previous pass could not tell them apart either. Added the test that was actually missing: `effort.agent_overrides`, which outranks the tier default, so the requested level genuinely reaches the renderer and `max` survives to `effort_rendered` end-to-end through the real CLI. Verified by probe before asserting. One test now pins the precedence rule and the other pins the #3007 behavior, and neither can be read as the other. That the clamp-visibility fields are what resolved this is a small argument for having added them. * chore(#3007): backfill changeset pr number to 3765 * test(#3007): put model-catalog under the mutation gate The Stryker shard showed as `skipping` on this PR despite the diff rewriting model-catalog's effort logic. That was legitimate, not a detection bug: `model-catalog` was never in scripts/mutation-matrix.cjs's COVERED map, so the whole module — including everything #3007 touches — sat entirely outside mutation scoring with has_work "false". Registered, with a dedicated spawn-free surface. tests/model-catalog.unit.test.cjs is new: 44 in-process tests, no runGsdTools, no child process, no filesystem, no temp dirs. That shape is not stylistic — it is the #2790 precedent this file already documents. Stryker's command runner treats a whole `node --test <file>` invocation as ONE test costing whatever its slowest case costs, and re-runs it per mutant, so pointing a shard at tests/model-resolver.test.cjs (which uses runGsdTools throughout) would reproduce exactly the 15-minute shard-cap cancellation #2790 hit. The integration file is unaffected and keeps running in full in the normal test job. Coverage spans the module rather than only the diff, because the score is measured over the whole file: effort rendering across every model and ladder level in both channels, the prototype-chain host guard, the exported enums and maps, isAnthropicFlavoredModel's provider namespacings, the profile projections, nextTier, and mergeEffortTierDefaults. The last two were nearly left out and are worth naming — every uncovered exported function is score given away, and mergeEffortTierDefaults turned out to have a genuinely interesting contract (#3531: a partial override merges over the built-ins rather than replacing them, and isValid gates the VALUE, not the tier name, so an unknown tier key is still merged in). Every expectation was probed against the built module before being asserted. minScore is 1 and that is a PLACEHOLDER, flagged as such in the registry comment. Floors in this repo are measured, not chosen — the existing entries sit at 94, 75 and 56 — and they can only be measured in CI, because mutation shards run `node --test`, which is hard-blocked locally. The first CI run on this branch reports the real number and the floor gets ratcheted to it before merge. A placeholder of 1 reaching `next` would make the gate decorative: it would pass whether or not a single mutant is ever killed. Note the target is "never regress from measured", not a fixed 80 — planning-inspect sits at 56 and is documented as an accepted ratchet candidate. * test(#3007): bootstrap model-catalog's mutation floor legally The placeholder floor was structurally illegal and the remote run said so. tests/mutation-matrix-ratchet.test.cjs guards the guard: every COVERED module must carry a matching RATCHET_BASELINE entry in the same diff, minScore must EQUAL that baseline, and it must be at least 50. `minScore: 1` failed all three. That is the ratchet working exactly as intended — a floor nobody can satisfy accidentally is the point of it. Bootstrapped at 50 in both places. Fifty is not a measured score and the comment says so plainly: it is the minimum the guard permits, and it coincides with Stryker's own configured `break` threshold, so it is the lowest legal starting point for a module that has never been measured. It still must be ratcheted to floor(measured) - 1 before this PR merges. Also corrected a real defect in the file's own instructions. "HOW TO UPDATE" step 1 read "Run the per-module Stryker shard locally" — which cannot be done here, and which the same file contradicts eighty lines further down, where the #2790 scores are recorded as "not a local run; mutation shards run `node --test`, hard-blocked in this repo's local environment". stryker.config.mjs confirms the command runner invokes `node --test` once per mutant, and .claude/hooks/block-local-node-test.sh denies exactly that. So the documented first step sends the next contributor at a wall. Rewritten to describe the path that works — push, read the measured score off the CI shard, then set the floor and its baseline together in one diff — and to say why local measurement is not available, so nobody rediscovers it the slow way. GOODHART SAFETY is untouched. The two-step is inherent to the environment rather than a shortcut: a floor cannot be measured before the first CI run exists, and the guard rightly refuses to accept an unmeasured one below its minimum. * test(#3007): ratchet model-catalog's mutation floor to its measured score The shard ran in CI and reported 59.62% — 248 mutants killed, 168 survived, no timeouts, no errors (run 32605073352, job 97108869486). Floor set to 58 per this file's own rule, minScore = floor(measured) - 1, which is the same arithmetic every sibling entry used: 57.03 to 56, 76.58 to 75, 95.65 to 94. Both halves moved together, because the ratchet guard asserts minScore equals its RATCHET_BASELINE entry and would reject them drifting apart. The spawn-free unit surface is vindicated by the clock: 57 seconds, against a 15-minute shard cap and a 9m46s frontmatter shard in the same run. That was the whole reason for creating tests/model-catalog.unit.test.cjs rather than pointing the shard at tests/model-resolver.test.cjs — #2790 recorded shards being CANCELLED at that cap when they targeted a runGsdTools-heavy integration file. The registry comment is rewritten rather than deleted. It previously warned that the floor was provisional and must not ship that way; leaving that text next to a measured floor would make the file lie in the other direction. It now records the measurement the way the sibling entries do, including that 59.62 sits below TARGET (80) and is therefore a ratchet candidate like planning-inspect at 56 — comfortably clear of its own floor with real room to grow. Raise it as the tests improve; never lower it. Worth stating plainly: 168 surviving mutants is not a clean bill of health. It is an honest floor for a module that had NO mutation coverage at all an hour ago, and it is now pinned so it cannot silently regress. --------- Co-authored-by: sim <sim@local>	2026-08-22 20:51:55 -04:00
Tom Boucher	8da2dd3ad2	feat(#2790): add read-only planning.inspect schema-v1 snapshot query (#3708) ... * feat(#2790): add read-only planning.inspect schema-v1 snapshot query Adds a read-only query emitting a schema-versioned JSON projection of .planning/ so downstream harness UIs can consume planning state without parsing GSD's Markdown a second time. Composed strictly from the ADR-3180 section 7 owners plus parsePlanDocument, parseRequirements and parseUatItems; markdown structure is read through the Markdown Sectionizer and Markdown Table Model seams. It declares its own flat external schema rather than serializing PlanningSnapshot, which is the diagnostic-rule subject and still growing. Extracts plan-document parsing out of cmdPhasePlanIndex into a shared leaf module so phase.plan-index and planning.inspect cannot drift, including the plan-id derivation both surfaces report. Also fixes parseRequirements dropping the separator delimiter used by the shipped requirements template, surfaced while wiring the requirement rows. * fix(#2790): close spec gaps and a raw-text test assertion found in review Review findings from the standards, spec and security passes: - phases[] rows carry goal and dependencies, the two per-phase elements the issue Summary names that had no corresponding field. Goal is bounded to the section's leading prose so the Depends-on line, the Plans checklist and the wave annotations are not duplicated into it. - requirement rows carry their own diagnostic codes, so a consumer no longer has to string-parse the global diagnostics subject to correlate. - roadmap_acceptance.checkbox is looked up through the phase-id key owners. It was compared raw against the on-disk directory name, so it read null for every real-world slugged phase directory and the evidence channel was inert. - the hostile-input test asserts the structured payload instead of matching the raw stdout string. The absence proof over raw stdout is kept deliberately. * fix(#2790): register planning in the runtime usage list and repair fixtures Remote runner reported 9 failures on 9b3f9aa. Two root causes, both fixed: - gsd-tools.cjs registered the planning family in HOST_COMMAND_ROUTERS but never added it to TOP_LEVEL_USAGE's Commands list. Those are two surfaces a parity test guards, and the top-of-file block comment is not the runtime help string. A real wiring gap that every local gate and three review passes missed. - the new suite's fixtures could not produce a resolvable phase set. STATE.md frontmatter omitted the milestone field, which ADR-3180 7.2 rule 1 makes the primary milestone selector, so the phase set scoped unscoped and every percentage was correctly withheld. Separately declarePhase returned a path without creating the directory, so a phase declared but never written to left phases empty. Both reproduced against the built module before fixing. No assertion was weakened. The withholding path is still exercised and still returns null when the roadmap is absent. * chore(#2790): backfill changeset pr number * test(#2790): cover every enumerated matrix row and contain a symlink escape Reverses a silent deferral. An earlier revision left 23 of the 78 enumerated matrix rows unimplemented and 7 more as one-off manual checks, with a paragraph in the artifact and the PR body describing the gap. CLAUDE.md is explicit that such a note is not a fix and is not surfacing. The rows are implemented instead and the manual-evidence bucket is gone: 49 test cases become 88, covering all 78. Writing the symlink row proved a real leak: a *-PLAN.md symlinked outside .planning/ had its content emitted into the payload, confirmed via a direct call and the spawned CLI. readDocument now resolves target and planning root with realpathSync and rejects an escape, returning the ordinary unreadable-document shape. Tested both ways, because a containment check that over-rejects is its own defect: an escaping symlink leaks nothing and degrades that plan alone, while a legitimately relocated .planning/ symlink stays fully readable. The three new modules are registered in the mutation COVERED registry, which had been reporting has_work false and skipping the Stryker gate entirely. Provisional non-binding floors so the shards run and report; raised to the measured value before merge, since the registry forbids calibrating from a local run. * fix(#2790): satisfy the mutation ratchet contract and scope the 1MB test Remote runner reported 16 failures on 8c451ed. Two causes. The COVERED registry has a paired contract the earlier commit violated: every module needs a matching RATCHET_BASELINE entry, and minScore must be between 50 and 100 with minScore === baseline. The provisional floor of 1 was illegal on both counts. All three modules now sit at 50 — the registry's own enforced minimum — with matching baselines. The score cannot be measured locally: the shard runs node --test, which this repo hard-blocks, so CI is the only source. Floors are raised to the measured value once this PR's shards report; a shard below 50 means the tests need strengthening, since the floor cannot go lower. The 1MB test was measuring the test harness rather than the product. The command handles the oversized payload correctly by spilling to a tmpfile and resolving it back, but the resolved stdout then exceeds runGsdTools' maxBuffer and the helper reports ENOBUFS. It now uses --pick so stdout stays one byte while the full 1MB document is still read and parsed end to end. * fix(#2790): wire containment across every document read this command drives An isolated security review of the containment control found the boundary logic sound but not comprehensively wired: two content reads reached the filesystem without it. An escaped phase DIRECTORY could enumerate external filenames into the file fields and diagnostic subjects. Both enumeration sites now containment-check the directory before reading. Worth recording that the leak was already prevented one layer earlier than the review claimed: Dirent#isDirectory() reports false for a directory symlink, so such a directory never becomes a phase row at all. The guard is defense-in-depth for a direct caller and for platforms where a reparse point reports as a directory. A *-VERIFICATION.md symlinked outside the root leaked one frontmatter value verbatim, because readVerificationStatus does its own read and copies an unrecognized status into the payload's next_action. Closed from the consumer side through that function's existing fs injection seam, so src/verification.cts keeps its signature and its other callers are untouched. The reviewer additionally rated a forged status: passed as an integrity bypass. It is not: anyone able to plant the symlink can plant a real VERIFICATION.md saying the same thing. The incremental risk is confidentiality, which is what these fixes close. src/plan-scan.cts is deliberately unchanged: isPlanSuperseded reads symlink-followed content but yields only a derived boolean, no document text. * test(#2790): give the mutation shards an in-process surface Two Stryker shards were CANCELLED at the 15-minute cap, not failed on score. CI log: 640 mutants instrumented, and the dry run reported 'Ran 1 tests in 20 seconds' because the shards pointed at the integration suite, where nearly every case spawns a gsd-tools subprocess and Stryker's command runner treats the whole test-runner invocation as a single test. 640 x 20s cannot finish in 15 minutes; at the kill it was 27/640 with an ETA over an hour. Every other COVERED module points at a property or unit file, and the workflow's own paths filter lists exactly those two patterns. In-process is the intended mutation surface; the shards were pointed at the wrong shape of test. Adds tests/planning-inspect.unit.test.cjs — 39 cases in 10 describes that spawn nothing and call the built modules directly. plan-document and the router need no filesystem at all, one being a pure content-to-object parser and the other taking an injected mock. The three shards now point here. The 91-case integration suite is untouched and still runs in the normal test job. * chore(#2790): ratchet mutation floors to the measured CI scores CI run 32392791843 measured all three shards, which is the only source the registry accepts — local runs count timeouts as kills and inflate badly. planning-command-router 95.65 -> floor 94 plan-document 76.58 -> floor 75 planning-inspect 57.03 -> floor 56 Applied the registry's own rule, floor(score) - 1, and updated RATCHET_BASELINE to match, since the ratchet test enforces equality. planning-inspect sits well below the file's target of 80 and is the obvious ratchet candidate as its tests improve. planning-command-router already exceeds the target. The placeholder comment about floors pending measurement is removed rather than left standing as a false statement. --------- Co-authored-by: sim <sim@local>	2026-08-20 13:42:43 -04:00
Tom Boucher	cbd180c5cd	test(#3147): bound the lint/changeset/docs cluster onto the process seam (#3181) ... * test(#3147): bound the lint/changeset/docs cluster onto the process seam Migrates 69 unbounded sync spawn sites across 24 files. Allowlist 73 to 49. Two shared helpers move: tests/helpers/graphify.cjs (6 importing suites) and tests/fixtures/index.cjs, whose three quoted-argument shell strings became single argv elements rather than whitespace splits. changeset-lint's throw-native git() helper routes to gitOrThrow; migrating it to bare runGit would have silently swallowed a failure that is loud today. ingest-docs goes the other way -- its catch never rethrew, it degraded failure into data every call site asserts on, so throwIfFailed would have thrown where the original returned. The design doc said otherwise and was corrected. tsconfig-noemit runs a real tsc --noEmit and takes a bespoke 180000ms per the ensure-runtime-build precedent, not the 30000ms build-hooks norm -- that norm is for a file copy, and sizing against a label rather than the work is the same error in the opposite direction. * test(#3147): add toLegacyResult and settle review findings The seam exposed a throwing adapter (throwIfFailed) but no non-throwing one, so eight files independently re-derived the same unwrap back to the legacy {status, stdout, stderr} shape. That is the third time this epic produced N copies of one mechanism -- seven throw wrappers in Wave 1, fifty-two timeout constants in Wave 2, eight result adapters here. The pattern is that whenever the seam does not expose a mechanism, every suite re-derives it. toLegacyResult now sits beside throwIfFailed, with its own tests. Two sites are deliberately NOT converted: changeset-cli's runRender and runRenderIn return {status, report, stderr} from parsed JSON and never a raw stdout, so they are a different shape family. lint-legacy-dir-name keeps its local GUARD_TIMEOUT_MS: 30000 matches the build norm numerically but bounds a lint probe, not hooks bundling, and importing it would encode a coincidence as a relationship. --------- Co-authored-by: sim <sim@local>	2026-08-07 15:45:51 -04:00
Tom Boucher	9ac0dfad58	chore(#2929): generalize prompt-budget into the shared context-composer seam (#2958) ... * test(#2929): capture prompt-budget parity corpus pre-refactor Phase 2 of epic #1671 generalizes prompt-budget's trim ladder into a shared context-composer seam. Its success condition is that review-prompt output does not change, and the only authority on "did not change" is the behavior that shipped before the refactor. Capture that behavior now, while it is still the live implementation. 47 characterization cases, every `expected` value computed by executing the current implementation rather than hand-authored — the independence CONTRIBUTING.md "Fixture provenance (#2371)" asks for. A corpus is only worth what it can detect, so this one was validated by mutation rather than assumed. Five deliberate defects were injected and each must be caught by at least one case: - the note reserve deducted unconditionally instead of only under pressure - the pressure test relaxed from `>` to `>=` - a no-op head-shrink still setting the shrunk flag - the per-plan floor dropped from the proportional share - drop order reversed Two of those exposed real holes in the first cut of this corpus, and the cases that close them exist because of it: - `>=` was caught by NOTHING. At exact cap the only trimmable fragment was a floored plan group, and the 1024-char floor absorbed the entire trim, so the mutation was byte-invisible. A3b/A3c put a droppable at exactly the cap, which makes the strict inequality observable as context kept vs omitted. - No case reached proportional-truncate at all — B6 and B7 both hard-failed the min-set pre-check first, leaving planTruncationPct at 0 across every case and the floor semantics entirely unexercised. Rebudgeted to 700 and 1100 so the min-set fits and the truncate step is actually reached; they now record 40.20% and 48.80%. The A4/A10 families sweep the pressure boundary from both sides, which is where this function has regressed before: CONTEXT.md's LEARNING.prompt-budget.boundary-gap records PR #3708 shipping two regressions that only fired when the baseline sat inside the NOTE_RESERVE_TOKENS band, because the suite paired a trivially-fitting budget with a trivially-overflowing one and never sampled between them. A4 pins that nothing is trimmed from the cap down to 81 tokens under it; A10 pins that pressure fires at +1. Together with A3b/A3c they satisfy row (d) of RULESET.TESTS.boundary-coverage.fixtures. Two facts the corpus establishes that the design notes had wrong: - "" and null sections are NOT distinguished. applyBudget uses truthy checks throughout, so an empty-string section is treated as absent: not rendered, not dropped, never recorded in `omitted`. B13b pins this while the ladder is actively trimming, where only the non-empty `research` is dropped. - Sizing matters. B12/B13 were first written at a budget where both hard-failed the min-set check and returned "", so comparing them compared two empty strings and proved nothing. Committed as its own commit, ahead of the refactor, and regenerated against the pre-refactor implementation, so the oracle is demonstrably independent of the change it will adjudicate. Refs #2929 * refactor(#2929): extract the context-composer seam from prompt-budget Epic #1671 needs prompt-budget's budget-trimming logic for a second consumer — per-runtime artifact emission — but it is walled inside the cross-AI review pipeline. Lift it into a shared seam so later phases can call it, without changing what the review pipeline emits. ADR-1671 specifies the composer as "priority + binary-search cutoff to a per-runtime budget". Read against the code it generalizes, that contract cannot express the thing being generalized. applyBudget is not a cutoff: it is a fixed five-step ladder in which each section carries its own shrink strategy, and only three of its eight sections are ever dropped. PROJECT.md is head-shrunk to N lines; plans are proportionally tail-truncated with a per-plan 1024-byte floor; instructions and roadmap are never touched at all. A cutoff composer sorts by priority and discards the tail — it has no way to say "shrink this one", "truncate that one but never below 1 KB each", or "these three are the only droppables, in this order". Building to the literal contract and routing prompt-budget through it would have silently changed review-prompt output, which is the one outcome this phase forbids. So shrink strategies are the core abstraction here, and cutoff becomes one strategy among them — the right one for per-runtime emission in Phases 3-4, not for this ladder. That is an elaboration of the ADR's intent, not a departure from it, and ADR-1671 is updated to say so. Three decisions worth stating: - The composer DECIDES; the caller RENDERS. composeWithinBudget returns a plan of surviving fragments and never a string. assemblePrompt's rendering is prompt-shaped (`## Roadmap`, `### <file>`, the note in position two), and owning it in the composer would force emission to adopt prompt-shaped rendering. The split is what lets one seam serve both consumers. - The budget unit is INJECTED via `measure(text)`. prompt-budget passes its chars/4 estimator; emission will pass a byte counter, which ADR-1671 requires for emission caps. The existing code converts a token budget to a character budget with a hardcoded `* 4`; that assumption is now an explicit `charsPerUnit` inverse, which is precisely what a byte unit needs in order to reuse this. - The entry point is `composeWithinBudget`, not `applyBudget`. That name already exists twice — src/prompt-budget.cts and src/graphify.cts, the latter being an unrelated graph-edge budget. A third would make every symbol search in this repo ambiguous, and it already misresolves: preflight and impact queries for "applyBudget" return graphify's. Behavior is unchanged and proven so: all 47 characterization cases reproduce byte-identically, and the corpus is mutation-validated rather than merely green (see the preceding commit). prompt-budget.cts drops from 436 to 343 lines and from eighteen mutable accumulators to two, both inside a helper copied verbatim. estimateTokens deliberately stays in prompt-budget and keeps its exact math: src/phase-estimation.cts re-exports it as measureTokens, and CONTEXT.md pins plan estimates and recorded actuals to that same scale, so moving or changing it would silently break the calibration loop. Refs #2929 * docs(#2929): document the context-composer seam and amend ADR-1671 Adds the INVENTORY row, the CONTEXT.md glossary entry (a PR gate for new domain modules), and a mutation-matrix entry for the new module. The ADR amendment is the substantive part. ADR-1671 specified the composer as "priority + binary-search cutoff to a per-runtime budget". Implementing Phase 2 established that a cutoff alone cannot express the function the platform generalizes, so the ADR now records shrink strategies as the core abstraction with cutoff as one strategy among them, reserved for per-runtime emission in Phases 3-4. Recording it in the ADR matters because Phases 3-6 are planned against that contract and would otherwise be planned against a mechanism that does not work. The mutation-matrix entry is not bookkeeping. Stryker scores per module against a named .cjs, so relocating the ladder out of prompt-budget.cjs would leave the extracted code unmeasured while prompt-budget's own score floated free of the logic it used to cover. context-composer gets its own entry at the same floor. Refs #2929 * test(#2929): pin the effectiveBudget rounding mode in the parity corpus An isolated correctness review found a real blind spot: mutating `Math.floor` to `Math.round` in the effectiveBudget calculation failed ZERO of the 47 corpus cases. Every (budget, safetyMarginPct) pair in the generator happened to produce a whole number, so floor, round and ceil all agreed and the rounding mode was entirely unpinned by a corpus whose whole job is to pin observable behavior. Three cases fix that by straddling the .5 boundary: A11 95 * 0.90 = 85.5 floor 85, round 86 -> the two disagree A12 97 * 0.90 = 87.3 floor and round agree; ceil (88) does not A13 93 * 0.85 = 79.05 same guard at a non-multiple-of-10 margin, so the margin arithmetic is exercised and not just the budget A11 alone catches the round mutation; all three catch ceil. Regenerated against the pre-refactor implementation (`git show 9557f8552:src/prompt-budget.cts`), so the expanded corpus keeps the independence property the original capture had. The corpus is now mutation-validated against seven injected defects, every one caught: unconditional note reserve, `>` relaxed to `>=`, no-op head-shrink setting its flag, the truncate floor ignored, drop order reversed, and both rounding-mode changes. Refs #2929 * feat(#2929): flexReserve floors and the byte-stable isolate prefix Two of issue #2929's "Done when" items were unimplemented rather than deferred, and an isolated review flagged them alongside my own audit. Both are part of ADR-1671's composer contract, so shipping the seam without them would have left Phases 3-4 building against a contract that does not exist yet. flexReserve is a per-fragment floor in measure units that every strategy must respect, which is what makes it different from the pre-existing floorChars: that one is a chars-denominated detail of proportional-truncate alone and is retained unchanged. A floored fragment is never dropped, is never head-shrunk below its floor, and raises its own proportional cap. A fragment already smaller than its floor is untouchable outright. Metadata gains `floored`, listing the ids whose floor actually prevented a trim — a guarantee no caller can observe is a guarantee no test can hold you to. isolate marks the byte-stable canonical prefix the ADR calls for: never trimmed, never dropped, but still counted, because a prefix excluded from accounting would silently under-count real context. Metadata gains `isolatePrefix` so a caller can hash or assert on the exact bytes. Declaring an isolate fragment after a non-isolate one throws: a prefix that is not at the front is not a prefix, and accepting it would make the cross-runtime stability claim meaningless. Adds tests/context-composer.test.cjs for the exact new semantics and tests/context-composer.property.test.cjs for the five invariants, including the budget-monotonicity property the issue names explicitly. Both are registered in the mutation matrix, since coverage does not migrate with relocated code. prompt-budget uses neither feature, and its output is unchanged: all 50 corpus cases still reproduce byte-identically. Refs #2929 * chore(#2929): allowlist the prompt-budget parity suite The parity corpus needs its own test file and that makes prompt-budget a three-file module against a limit of two. The lint offers consolidation or an allowlist entry with justification; the entry is the right call here. Consolidation would mean folding the characterization suite into prompt-budget.test.cjs, which is the one thing that should not happen to it. The parity suite is a distinct concern with a distinct lifecycle: it is generated rather than hand-written, it is named by scripts/mutation-matrix.cjs as its own scoring target, and its failure means something categorically different from a unit-test failure — not "this behavior is wrong" but "observable output moved". Burying it inside a general unit file would obscure exactly that signal. The allowlist is an identity ratchet, so this entry pins today's three exact filenames: adding a fourth still fails, and dropping back to two requires removing the entry. Refs #2929 * fix(#2929): register the new module with two gates it was missing The remote matrix caught three defects that no local check could, because the local runner is blocked in this repo and these suites had therefore never executed. Eight failures, identical on node22 and node24, so nothing environment-shaped. Two are the new-module ripple. A net-new src/*.cts lands in six places and this change had reached four of them — .gitignore, INVENTORY, the manifest, and the CONTEXT.md glossary — while missing the ESLint ignore list (tsc OUTPUTS must not be linted; repo-invariants asserts linted-xor-ignored) and the mutation ratchet baseline (a deliberate review-visible mirror of the matrix floors, which every COVERED module must carry). Both are now registered, the ratchet at the same floor of 66 the matrix declares. The third was a test asserting an outcome it had made impossible. It set budget:1 alongside a 400-char required fragment, so the group budget came out at -99 and the proportional-truncate step was skipped entirely — the deliberate "non-positive group budget is skipped, never clamped" rule inherited from the original ladder. Nothing was trimmed, and the test then asserted a truncation. Rebudgeted so the step actually runs, with the arithmetic written out in a comment so the next reader does not have to re-derive why 120 rather than 80. Fixing that surfaced a genuine bug in the composer. `floored` is documented as recording fragments whose flexReserve prevented a trim that would otherwise have happened, but the push sat in the else-branch of "content did not change", so it only fired when nothing was trimmed at all. A fragment truncated to a reserve-raised cap has also had a trim prevented — 40 characters' worth in the test above — and was silently absent from the field that exists to make the guarantee observable. The condition was already right; it was in the wrong branch. Now recorded on both paths: a drop prevented outright, and a truncation capped higher than the share alone would have allowed. Parity is unaffected — prompt-budget never sets flexReserve, so the branch is unreachable from every corpus path, and all 50 cases still match. Refs #2929 * chore(#2929): backfill changeset PR number (#2958) * chore(#2929): correct the corpus case count in the changeset fragment --------- Co-authored-by: sim <sim@local>	2026-07-31 23:03:13 -04:00
Tom Boucher	98866a0c69	feat(#1187): per-module Stryker mutation-score ratchet (ADR-456 80% floor) (#1200) ... * feat(#1187): per-module mutation-score ratchet + graduate core-utils ADR-456's 80% mutation floor was unenforceable as a single global break=50: 4 of 6 covered modules sit at 63-79% and forcing them to 80 would require brittle exact-string assertions on equivalent string-literal mutants (a Goodhart's-Law trap). Instead, each covered module declares a minScore floor (locked at its measured score, TARGET 80) enforced per CI shard via stryker --break, ratcheting up over time without brittle tests. - mutation-matrix.cjs: minScore per module + TARGET_MUTATION_SCORE=80, emitted in the matrix; require.main guard + exports for testability. - mutation.yml: per-shard --break <minScore>. - stryker.config.mjs: global break 50->60 as a local backstop (CI uses minScore). - Graduated core-utils (measured 77.5%, floor 75). - context-utilization 79.5->92.3% via behavioral killers (state classification outputs + error-value contract, not exact-string matches) -> minScore 80 (TARGET). - ratchet-integrity guard test (28 cases). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix(#1187): pass mutation break via MUTATION_BREAK env (no stryker --break flag) Adversarial review caught that Stryker 9.x has no --break CLI flag, so the per-shard 'stryker run --break <minScore>' errored out every mutation shard. Read the per-module floor from process.env.MUTATION_BREAK in stryker.config.mjs and set it per shard via env in mutation.yml. Red-green verified: MUTATION_BREAK=99 exits 1, =80 exits 0. Also make the ratchet guard monotonic (RATCHET_BASELINE floors; lowering a floor now fails the guard unless the baseline is edited). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix(#1187): fail closed on bad MUTATION_BREAK + monotonic ratchet baseline Code review: Number(env)||60 failed OPEN — an empty/invalid MUTATION_BREAK (e.g. a future module missing minScore -> matrix expands to '') silently degraded the shard to break 60, letting a high-floor module regress undetected. resolveMutationBreak() now returns 60 only when the env is truly unset (local backstop) and THROWS on present-but-empty/non-numeric/out-of-range (fail closed); stryker.config.mjs imports it via createRequire. Also make RATCHET_BASELINE an equality mirror (=== not >=) so any floor change is explicit in review and no floor can be silently lowered. Tests: 46 (incl resolveMutationBreak cases). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix(#1187): recalibrate config-schema/prompt-budget floors to CI scores First CI mutation run failed two shards: the floors were set from local Stryker runs whose TIMEOUTS were counted as kills (env-variable), inflating scores. CI runs with timeout~0, so the real deterministic scores are lower: - config-schema: local 69.7% -> CI 54.55% (5 local timeouts vanished) -> floor 52 - prompt-budget: local 99.6% -> CI 68.33% (239 local timeouts vanished) -> floor 66 Calibrate floors from CI (the documented source of truth) and record the lesson in the comment so future floors aren't set from timeout-inflated local runs. Baseline updated to match. The other 5 shards passed (deterministic CI scores above their floors). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>	2026-06-14 08:25:41 -04:00