RESTAMP: 1.91.279.0 -> 1.91.280.0 (stamped; queue_mode=stamp-at-merge base=origin/main@b3e94a3f8d1c policy=base origin/main@b3e94a3f8d1c:.gstack/ship-policy.json)

STAMP_VERSION: 1.91.280.0
stage version: done (wrote 1.91.280.0)
stage release_tool: done (bun scripts/gen-agents-digest.ts)
stage changelog: done (Unreleased -> 1.91.280.0)
stage stamp_paths: skipped (no tentative mention newer than the fork version 1.91.277.0 and none of base 1.91.279.0 on a line this branch added)
stage mirror: skipped (no mirror in policy)
stage apply: done (4 file(s))
FILES: CHANGELOG.md VERSION agents-digest/gstack-AGENTS.md package.json
MERGE_CONDITION: sha=a755d3155d0921b05b9ca350710ef6724658376c
GSTACK_RESULT: skill=restamp status=complete run=/workspace/gstack/.gstack/tmp/restamp-journal.json

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

CHANGELOG.md

@@ -1,6 +1,6 @@
 # Changelog
 
-## [Unreleased]
+## [1.91.280.0] - 2026-10-27
 
 **Three free-tier tests measured something other than what they assert, and the Ubicloud quota wait shaved its own budget.** The daily flaky pass (GSTA-128) traced the free tier's open timing reds to measurement defects, not product ones: `quota_preflight` in `scripts/ubicloud/ubi-runner.sh` pinned its deadline to the wall clock before the first usage call, so a slow CLI spawn on a loaded runner printed "waiting up to 2s more" where the test expects the full 3 s; the pregate-remote teardown test read the `.down` file as soon as it existed, before the stub's `echo` had written the line; and the design-lint linearity check timed one wall-clock run of the 8000-label lint against a 500 ms floor on a shared runner. Under eight busy loops on four cores the quota case went from 8 of 12 green to 12 of 12. No ms bound, threshold, trial count or retry changed.
 
@@ -15,6 +15,51 @@
 ### Gate integrity
 - Four hunks listed, three read, none a relaxation. The teardown wait is stricter, the watchdog assertion is unchanged with a message added, and the linearity measure keeps its bound (measurement above).
 
+## [1.91.279.0] - 2026-10-27
+
+**A licence line or a changelog entry no longer buys a full paid gate, the review loop stops re-reporting findings the owners have already ruled below the floor, and three follow-ups from the Blast radius ship land.** #3499: every PR that touched `NOTICE.md`, `LICENSE`, `licenses/` or `ETHOS.md` restored every gate case and judge because no paid case was mapped to them (#3493's docs-only PR planned 82 cases, about $30, for a notice and a changelog entry). #3504 and #3485: the same classes of finding (a schema lagging its writer, a helper nobody calls, a re-dispatch, a refusal one step late) reached the owners again on every ship after being judged below the floor once. GSTA-121's ship left three gaps. The Captain's merge-ask template and `docs/skills.md` lacked the `## Blast radius` item the PR body now carries, an 11-digit GitHub Actions run id after a run word was redacted from an approval record as a phone number, and the remote free gate failed with `PREGATE_REMOTE_INFRA` when a reviewer ran git in the checkout while the gate tarred `.git`.
+
+### Added
+- **`RELEASE_METADATA_FILES` in `scripts/test-pr-profile.ts`** (`NOTICE.md`, `LICENSE`, `licenses/**`): release metadata no paid case reads, exempt like `CHANGELOG.md` and `VERSION` already were in `knownNonBehaviorFile`; a PR of only these files selects no paid case and keeps mode `pr`; `captain/templates/**` (the crew templates no paid case reads, found as the one unknown dependency of this PR's own selection) joins `FREE_ONLY_PR_FILES`. `ETHOS.md` is registered under the cases and judge of the two prompts that read it (`/office-hours`, `/plan-ceo-review`) in `test/helpers/touchfiles-data.ts`, so an ETHOS edit selects those cases (three gate cases and one judge today) instead of the whole gate; `SECURITY.md` replaces it as the tests' example of an unregistered file. The "unknown dependency restores the full gate" rule itself is unchanged.
+- **`docs/review-policy.md#rulings-below-the-floor`**: six finding classes put to the floor by #3485 and #3504 and ruled below it (own-artifact hardening, unwired or unread, bounded repeat work, conservative refusal, artifact presentation, test gap), each with what it looks like and why it fails the floor; a finding in one is a note named by its ruling from pass 1 on. The review army's lead judgment (`scripts/resolvers/review-army.ts`, rendered into both `review-army.md` sections) gains the one-line **On record** ground that applies the table; its other grounds were tightened by the same byte count so the generated review SKILL.md stays at its pins. The floor ("would fail a test, break production, or lose data or secrets") is unchanged.
+- **`docs/TESTING_INTERNALS.md` Windows free lane**: the command that shows whether one test is in the curated Windows subset and why not, and which list (`KNOWN_WINDOWS_SAFE`, `WINDOWS_PROBE_SAFE`, `KNOWN_WINDOWS_INCOMPATIBLE`) moves it.
+- **`docs/skills.md` `/ship` "The PR body"**: the four Capy sections in order (What changed, Why, Blast radius, Approvals), how to write the Blast radius item, and what `gstack-prose-lint` gates. `captain/templates/merge-ask.md` carries `## Blast radius` between Why and Approvals; `bin/gstack-pr-shots`'s example caption and the `capy-pr` section heading name it too.
+
+### Fixed
+- **`lib/redact-patterns.ts` `isCiRunIdentifier`**: an 11-digit GitHub Actions run id right after a run word on the same line (`screen 40477638211 GREEN`, `gate=…`, `rerun passed …`, `run #…`) is not `pii.phone.e164`, in every scanner (`gstack-approvals add`, `gstack-decision-trail`, the pregate secrets check, `gstack-post`); the run-word regex moved there from `lib/gstack-post.ts`, which imports it for its chained-id allowlist. A 10-digit national number after a run word, a bare 11-digit number, an E.164 `+` number and `job`, `id`, `workflow` or `attempt` before one still report; a national number written without its `+` right after a run word reads as a run id, the accepted residual of the ruling that a run id is not a phone number (`gstack-post` has read it that way since GSTA-121).
+- **`scripts/ubicloud/ubi-runner.sh pack`**: the live `.git` is never tarred. The refs are fetched into a fresh repository through git's transport (`--update-shallow`, so a shallow CI or Capy checkout packs), the `remote.*` and `branch.*` config is replayed so `origin/<base>` and each branch's upstream resolve on the VM, a detached HEAD gets a `detached-<sha12>` branch, the index is copied with the blobs staged against HEAD (a VM `gstack-wtree` can read a staged file), and a pack that fails removes its temporary repository; a review subagent's `git diff` or a trial merge's `worktree add` during the first minute of a remote gate no longer makes tar exit 1 (`PREGATE_REMOTE_INFRA`). The linked-worktree path uses the same snapshot.
+- **`lib/pregate/import-graph.ts`**: a relative string literal (`'../ship/sections/pr-body.md.tmpl'` through `new URL(…, import.meta.url)` or `readFileSync`) is a dependency edge from the file that names it, like a relative import; `test/ship-publication-gates.test.ts` is now selected when the pr-body template changes (the remote free gate missed it in GSTA-121's ship). A literal that climbs out of the repository names nothing.
+- **`scripts/lib/windows-curation.ts`**: `test/pregate-remote-stamp-only.test.ts` is `KNOWN_WINDOWS_INCOMPATIBLE` (it expects the Linux-host lanes verdict and skips itself on win32), as #3485's follow-up asked.
+
+### Tests
+- `test/paid-free-boundary.test.ts`: release metadata alone and beside `CHANGELOG.md` keeps mode `pr` with no paid case; `ETHOS.md` with `NOTICE.md` selects the two skills' gate cases and judge and nothing else. `test/pr-dependencies.test.ts`, `test/paid-pr-profile.test.ts`, `test/exercised-map.test.ts`: `SECURITY.md` as the unregistered example.
+- `test/redact-context-exemptions.test.ts`: seven run-word rows with run 40477638211 and five controls; `test/approval-record.test.ts`: an approval whose words cite `screen 40477638211 GREEN` is recorded verbatim, a phone number still refuses the row; `test/pregate-allowlist.test.ts`: its run-id fixture reads `job <id>`, which the engine does not exempt, so the allowlist mechanism it tests still has a finding to silence.
+- `test/ubicloud-runner.test.ts`: pack of a plain checkout under twenty seconds of index, worktree and ref churn, three times; the unpacked snapshot has the same HEAD, `origin/main`, remote URL, status and `gstack-wtree` fingerprint, a detached HEAD lands on `detached-<sha12>`, a pack with no index exits 1 and leaves nothing in `TMPDIR`, and a branch's upstream survives the snapshot.
+- `test/pregate-import-graph-paths.test.ts`: relative literals through `new URL` and `readFileSync`, one and two directories up, and a literal that escapes the repository.
+
+### Gate integrity
+- No threshold, timeout, trial count or case kind changed. The paid selection narrows only for `NOTICE.md`, `LICENSE`, `licenses/**`, `captain/templates/**` and `ETHOS.md`; every other unmapped file still restores the full gate. Generated SKILL.md sizes stay at their pins and the ship parity union stays under its 1.7500 cap; no allowlist or cap moved.
+
+## [1.91.278.0] - 2026-10-27
+
+**The merge train's tooling stops failing the lanes for things the tooling itself could see.** A head whose only conflict with main was the CHANGELOG got its Free Tests through the dispatch route and the paid screen still called it NOT-RUN; the receipt-versus-prose check read the canonical `REVIEW_LOOP` line as a false mismatch; a PR that writes a new `*_KEY_VERSION` constant as a tentative next-minor version (the convention #3511 introduces) could not clear the train's stamp-only receipt; a pooled pin learned that main had moved onto its files only when the train was built; a watcher had no way to leave a stuck non-required check out of its verdict; and a lane that passed the slot call's remaining order to `gstack-restamp` without its own PR got the same stale list back as the fix. Backlog items from GSTA-119, #3485, #3495 and the 2026-10-27 slot calls.
+
+### Fixed
+- **`scripts/eval-readiness.ts`** (#3508): readiness accepts a `workflow_dispatch` Free Tests run at the exact head beside GitHub's `pull_request` run, so a head the dispatch route of docs/ship-policy.md tested starts its paid screen. GitHub's `pull_request` run decides whenever one exists (it built the merge ref; a dispatch builds the head alone, so a newer green dispatch never overrides a red `pull_request` run and an extra red or running dispatch never holds a head its `pull_request` CI cleared); the newest dispatched run decides only for a head with none, a red one blocks, and a `push` run still counts for nothing.
+- **`lib/ship-receipt-prose.ts`** (#3513): a number that is the value of a `key=` or `key:` pair is that key's claim alone, whatever the whitespace around the separator, so `REVIEW_LOOP: passes=2 blocking_found=1 deferred=0 noted=4 fixed=8 scope=frozen` and `passes: 2 blocking_found: 1` parse as five claims that all match the receipt instead of a `2 blocking_found` claim.
+- **Stamp paths under the next-minor key convention** (#3512): the tree receipt (`lib/tree-receipt.ts`), the carry (`lib/ship-receipt.ts` `stampOnlyText`) and the train's tree-modulo-stamps digest (`scripts/train-gate.ts` `stampFreeTree`) accept a stamp_paths line whose stamp-form version moved to the released version whatever it was before, not only when it was the base's version. The receipt keeps the stamp_paths stage's own bounds (`lib/restamp.ts` `stampOnlyLine`): a moved version older than the gated head's is a released key moved by hand, and one of another width (a 3-digit pin in a 4-digit repository) is never a stamp candidate, so both stay `changed`. The digest collapses every stamp-form mention, so a tentative key and a base-version mention on a base-side line digest the same before and after the slot stamp and `train-backstop` finds the record that covers main.
+- **`gstack-restamp --expect-order`**: the `QUEUE_STALE` refusal prints the corrected command instead of the stale list: this PR first when the list left it out (with the message saying every entry read as a predecessor), PRs merged into the local base dropped, an unmerged predecessor kept ahead with a note that it must merge first, and a predecessor GitHub calls merged that the local base does not carry yet kept ahead too, with `fix: git fetch origin <base>` (dropping it would let the recompute, pinned to that local base, allocate the version it already took). The base-moved refusal carries `--pr` too.
+
+### Added
+- **`gstack-ship-queue ready` and `train build` read gstack-pr-wait's base-overlap event from git** (handoff GSTA-119; the crew template's sibling trial merge, done by the tool). `ready` prints `BASE_OVERLAP: <base> moved to <sha7> and touches files in this PR: <files>; the train's merge onto <base> is clean` (or the conflicting files) when the base moved onto the PR's files since it forked, and under `--no-screen` a conflict is `QUEUE_TRAIN_CONFLICT` at READY rather than at the train, with the template's fix (report it to the controller; merge the base on its word). A git read that fails prints `conflicts unknown` and refuses nothing. `train build` runs the scratch merges before the fill rule. A seated pin whose files the base moved onto gets `#<pr> BASE_OVERLAP: <base> moved to <sha7> and touches files in this PR: <files>` (the chain's merge onto the pin's own prefix is the verdict), a pin that does not merge is `QUEUE_TRAIN_CONFLICT` while the train is still filling, and a clean pool prints `MERGE_CHECK: T1..T<k> merge clean onto <base>@<sha>` ahead of `QUEUE_TRAIN_WAITING`, so the waiting verdict is also a dry-run merge (#3485).
+- **`gstack-pr-wait --exclude <check,...>`** (#3495 follow-up): a named non-required check is left out of the verdict as `gstack-ci-gate --exclude` does, so it neither holds `--until green` while pending nor blocks when it fails; a required check cannot be excluded and the first poll logs the refusal, and while a required-check source cannot be read nothing is excluded (the set may be short), as `gstack-ci-gate` refuses its overrides.
+
+### Tests
+- `test/eval-readiness.test.ts`: a dispatched-only head decides green, red and running, a `pull_request` run outranks a dispatched one either way; the runs listing is no longer narrowed to `pull_request`. `test/ship-receipt-prose.test.ts`: the exact `REVIEW_LOOP` line with `passes=2 blocking_found=1` and with `passes=1 blocking_found=0`, the `key: value` spelling with whitespace, plus a real disagreement on the same line. `test/tree-receipt-stamp-sensitive.test.ts`, `test/train-gate.test.ts`, `test/gstack-ship-receipt-docs.test.ts`: a tentative key moved by the restamp is a stamp in the receipt, the digest and the carry; a renamed key, a released key moved by hand or a mention of another width is not. `test/gstack-restamp-queue.test.ts`: the corrected `--expect-order` in each refusal, the own-PR-first lesson and a merged predecessor the local base does not carry. `test/pr-wait.test.ts`: `--exclude` parsing, pending and failed excluded checks, a refused required name, an unreadable required set, and the flag through `prWaitMain`. `test/ship-queue-train-ops.test.ts` and `test/ship-queue-train.test.ts`: the `BASE_OVERLAP:` lines at READY and at build, the `--no-screen` conflict refusal, and `MERGE_CHECK:` beside `QUEUE_TRAIN_WAITING`.
+
+### Gate integrity
+- No threshold, timeout, trial count or case kind changed. No skill prose changed.
+
 ## [1.91.277.0] - 2026-10-27
 
 **office-hours stops carrying its own copy of the shared AskUserQuestion subsections.** After the shared-preamble hoist, office-hours still inlined whole copies of the shared `### When AskUserQuestion is unavailable or a call fails` and `### Format` subsections, because each held one office-hours addition: the `Q<N>` open-question form. A session's first office-hours invocation read those subsections twice, once from the shared file and once from SKILL.md. The `Q<N>` form and its routing sentence now render as office-hours' own `### Open-ended questions` subsection at the end of the AskUserQuestion section, so the shared subsections apply as they are and office-hours' SKILL.md drops 6,328 bytes (its parity union goes from 160,952 to 154,624, under the pre-hoist 1.18 cap).

VERSION

@@ -1 +1 @@
-1.91.277.0
+1.91.280.0

agents-digest/gstack-AGENTS.md

@@ -1,4 +1,4 @@
-# gstack digest v1.91.277.0 — regenerate/re-copy after upgrading gstack
+# gstack digest v1.91.280.0 — regenerate/re-copy after upgrading gstack
 
 Behavioral rules from gstack (https://github.com/garrytan/gstack), compressed
 for agent hosts without a full skill install. The full skills add workflows,

bin/gstack-pr-shots

@@ -33,7 +33,7 @@ const HELP = `gstack-pr-shots — before/after image of a change for a PR body (
   [--title <text>] [--name <slug>] [--out <dir>] [--json]
 
 Example:
-  gstack-pr-shots --path ship/SKILL.md --heading "Step 19: Create PR/MR" --caption "PR body opens with What changed, Why and Approvals under Capy"
+  gstack-pr-shots --path ship/SKILL.md --heading "Step 19: Create PR/MR" --caption "PR body opens with What changed, Why, Blast radius and Approvals under Capy"
 
 ${exitTableHelp()}`;
 

bin/gstack-pr-wait

@@ -2,7 +2,7 @@
 /**
  * gstack-pr-wait — watch one pull request and print one line per real event.
  *
- *   gstack-pr-wait <pr> [--repo owner/name] [--until green|ready|any-change] [--interval SECONDS] [--timeout MINUTES]
+ *   gstack-pr-wait <pr> [--repo owner/name] [--until green|ready|any-change] [--interval SECONDS] [--timeout MINUTES] [--exclude CHECK,...]
  *
  * Exit 0 when the --until condition is met or the PR merged, 1 when it needs
  * action (PR_WAIT_BLOCKED), 2 on bad input, a duplicate watcher or a local

bin/gstack-ship-receipt

@@ -446,7 +446,7 @@ function cmdCarry(rest: string[], json: boolean): number {
   const oldVersion = versionAt(from);
   const newVersion = versionAt(head);
   const allMetadataGlobs = [...CARRY_METADATA_GLOBS, ...extra.metadata];
-  const changed = changedRaw.split('\n').filter(Boolean).map(p => ({ path: p, class: classifyCarryPath(p, extra, matchesAny(allMetadataGlobs, p) ? { before: blobAt(from, p), after: blobAt(head, p), oldVersion, newVersion } : undefined) }));
+  const changed = changedRaw.split('\n').filter(Boolean).map(p => ({ path: p, class: classifyCarryPath(p, extra, matchesAny(allMetadataGlobs, p) ? { before: blobAt(from, p), after: blobAt(head, p), oldVersion, newVersion, templates: policy?.stamp_patterns ?? [] } : undefined) }));
   // Pathspecs by class: tests and metadata are excluded from the product patch; the test patch is the tests alone.
   // A metadata-listed path whose change was not a stamp stays in the product patch: every glob that matches it is dropped from the exclusion.
   const excludeGlobs = (globs: readonly string[]) => globs.map(g => `:(exclude,glob)${g}`);

captain/templates/merge-ask.md

@@ -10,7 +10,10 @@ it before training the PR.
 <one to three images, highest-signal first; a collapsed text version under each>
 
 ## Why
-<the cause and the user-visible effect, two or three sentences>
+<the problem in the person's words first, then the cause and the user-visible effect, two or three sentences>
+
+## Blast radius
+<what else the change touches and the one fact that makes that safe, with the evidence-ladder rung reached>
 
 ## Approvals
 <the table from gstack-approvals render>

capy/skill/gstack/captain.md

@@ -126,9 +126,12 @@ Every PR you send READY on, and the READY report itself, has:
    With no browser that line says `unavailable`: screenshot the printed `shot.html` with your own browser.
    Upload each PNG with Capy's PR asset tool (`tools.pr.uploadAsset`), only for public repositories, and embed
    the URL it returns with the caption as alt text.
-2. **Why.** The cause and the effect the user will see, in two or three sentences.
-3. **Approvals.** `gstack-approvals render`.
-4. **Evidence.** Checks green at the head, gate or train results, flaky cases.
+2. **Why.** The problem in the person's words first, then the cause and the effect the user will see,
+   in two or three sentences.
+3. **Blast radius.** What else the change touches and the one fact that makes that safe, with the
+   evidence-ladder rung that fact reached.
+4. **Approvals.** `gstack-approvals render`.
+5. **Evidence.** Checks green at the head, gate or train results, flaky cases.
 
 The READY report attaches the key image (Capy message attachment), so the Captain can show it inline.
 /ship does all of this when `GSTACK_CAPY=1` (its `capy-pr` section). Outside Capy, the PR body is unchanged.

docs/TESTING_INTERNALS.md

@@ -1097,8 +1097,10 @@ PR selection (`scripts/test-pr-profile.ts`) classifies every changed file. A fil
 mapped by a paid case's touchfiles selects that case. These files are exempt and
 select nothing: docs, any non-paid `*.test.ts(x)`, `FREE_ONLY_PR_FILES` (the
 free and paid duration seeds, `scripts/ubicloud/**`, every workflow that runs no
-paid evals, and free-only tools), and the fixtures in `FREE_FIXTURES`
-(`test/helpers/free-fixtures-data.ts`, each fixture with its free consumers).
+paid evals, `captain/templates/**`, and free-only tools), the release metadata
+in `RELEASE_METADATA_FILES` (`NOTICE.md`, `LICENSE`, `licenses/**`), and the
+fixtures in `FREE_FIXTURES` (`test/helpers/free-fixtures-data.ts`, each fixture
+with its free consumers).
 Unmapped files under `DERIVABLE_PREFIXES` (bin/, lib/, scripts/, browse/,
 design/, make-pdf/, hosts/, extension/, model-overlays/, agents/, test/helpers/,
 capy/) are not unknown: `scripts/pr-dependencies.ts` derives the cases whose
@@ -1236,7 +1238,12 @@ GitHub-hosted restore never sees (until GSTA-48 every restore missed).
 **Readiness (R1).** Paid slices for a PR head start only when the PR has been
 quiet for the window (repository variable `EVALS_QUIET_MINUTES`, default 10,
 measured from the run's creation) and Free Tests has concluded success on the
-same head SHA. The `readiness` job (`scripts/eval-readiness.ts`) sleeps out the
+same head SHA, as GitHub's `pull_request` run or, for a head with none, as a
+`workflow_dispatch` run on the branch (the dispatch route of docs/ship-policy.md
+for a head whose only conflict with the base is the CHANGELOG, which gets no
+`pull_request` CI; the `pull_request` run built the merge ref, so it decides
+whenever it exists, a dispatched run decides only without one, and a red
+dispatched run blocks like any other, #3508). The `readiness` job (`scripts/eval-readiness.ts`) sleeps out the
 window with no API calls, since a newer push cancels the run anyway, then polls
 Free Tests every minute for up to 80 minutes. Free Tests red on the head fails
 the run without paid work; re-run Free Tests, then re-run this workflow (a
@@ -1501,6 +1508,20 @@ summary). Files a probe proved safe despite a `/tmp` literal are force-included
 `WINDOWS_CURATED_FLOOR`; when Windows-safe tests are deleted on purpose, lower it
 in the same PR with a reason.
 
+To see whether one test is in the subset, and why not when it is excluded:
+
+```bash
+bun run scripts/test-free-shards.ts --windows-only --list | grep test/<name>.test.ts
+```
+
+A test a content pattern excludes wrongly joins the subset through
+`KNOWN_WINDOWS_SAFE` (with its reason) or, for a `/tmp` literal a probe proved
+safe, `WINDOWS_PROBE_SAFE`. A test that fails on `windows-latest` for a reason
+the patterns cannot see (it expects the Linux-host lanes verdict, skips itself
+on win32, drives a POSIX signal path) leaves it through a
+`KNOWN_WINDOWS_INCOMPATIBLE` entry naming the reason; rerun `--list` to confirm
+the file moved.
+
 Native qualification campaigns (Windows cookie extraction, Windows launch
 diagnostics, Dia on macOS) run only on dispatch of `native-qualification.yml`:
 

docs/captain.md

@@ -232,7 +232,7 @@ recommendation at full scope. `gstack-owner-brief render --captain` prints it fr
 autoplan run.
 
 **C. The merge ask** ([template](../captain/templates/merge-ask.md)): captioned
-before/after images, why, the Approvals table, and the evidence. /ship writes the PR
+before/after images, why, blast radius, the Approvals table, and the evidence. /ship writes the PR
 body in this shape on Capy; the controller's merge report shows the key image inline.
 
 ## The digest

docs/pregate.md

@@ -456,7 +456,9 @@ receipt for H when the two trees differ only as a stamp does:
   that does not account for every path `git diff-tree` lists between the two
   trees. The version rule is the tree
   receipt's: a `stamp_files`, `stamp_paths` or `release_outputs` hunk is a stamp
-  only as the rewrite from the gated tree's version to the head's (so two trees
+  only as the rewrite to the head's version from the gated tree's version, or
+  from a newer tentative version of the repo's width that the stamp moved (a
+  `*_KEY_VERSION` written as the next minor, #3512) (so two trees
   at the same version refuse such a hunk as `other`), while the built-in
   version-file, manifest, lockfile and CHANGELOG rules compare content and do
   not check the direction of the version move. The obligation stays uncleared; rerun
@@ -564,7 +566,10 @@ that reads them ships. The worked gbrain configuration is in
 
 The `secrets` check cannot be downgraded, and that is right for a key. It also
 refused 1,800 per-trial dummy tokens in eval receipts (`"key": "sk-trial-<id>"`
-reads as `env.kv`), an 11-digit Actions run id read as a phone number, and
+reads as `env.kv`), an 11-digit Actions run id read as a phone number (one
+after a run word such as `run`, `screen` or `gate` is exempt in the scanner
+since the GSTA-121 follow-up; the allowlist still covers one after any other
+word), and
 excalidraw.com's public Firebase key that every rebuild of a bundle re-adds
 (harness-evals #81/#83, GSTA-102, GSTA-106). `allowlist` in the policy names
 those, each with its scope and its reason:

docs/review-policy.md

@@ -39,6 +39,31 @@ floor above; everything else is INFORMATIONAL. Data loss, a trust-boundary
 bypass and uncapped spend stay forced CRITICAL (the sideways sweep rule); no
 lane, specialist, dedup or saved decision downgrades them.
 
+<a id="rulings-below-the-floor"></a>
+### Rulings below the floor
+
+The floor is a test, and these classes have been put to it, each more than
+once, and failed it (#3485 from the /ship of #3484, #3504 from the /ship of
+#3487). A finding in one of them is a **note** from the first pass on, named
+by its ruling in the PR body's `## Review notes`; it never earns a fix round
+or another review pass, and the lead judgment stage drops it from the defect
+count with the pointer. The floor itself does not move, so a finding that also
+loses data, crosses a trust boundary or spends without a cap stays CRITICAL
+whatever class it resembles.
+
+| Ruling | What it looks like | Why it is below the floor |
+|---|---|---|
+| **own-artifact hardening** | A check missing between two things the repository's own tooling writes: a screen verdict not compared with the receipt it belongs to; a receipt schema that does not declare a key the writer records; a policy loader that defaults when the base branch's own file is invalid; a size read that follows a symlink inside the repository. | The only writer is the repository's maintainers through its tools; no caller inside the intended case reaches the gap, and no test fails. A writer outside that trust boundary makes it a bypass, and CRITICAL. |
+| **unwired or unread** | A helper with no caller, a field written and never read, a duplicated helper, a header comment naming a step that does not run it, a key-version constant that moved before any release carried the old value. | Nothing takes the path, so nothing breaks; it is dead code to delete or wire in its own change. |
+| **bounded repeat work** | A re-dispatch on every call, a redundant screen after main moves, an orphaned scratch branch, a refusal that could come one step earlier in the sequence, a settle that could wait one more check. | The operation completes, nothing is lost, and the cost is one more bounded run; uncapped spend is the sideways sweep's CRITICAL, not this class. |
+| **conservative refusal** | Stale state scoped wider than its cause: a RED verdict that keeps refusing a head after main moved, a pending result that keeps refusing after it settled. | A refusal costs a rerun and never passes something wrong; a wrong pass would be the floor. |
+| **artifact presentation** | A string from an artifact printed verbatim into a parsed line (whitespace, control characters); a classifier that reads a non-zero exit code beside a non-RED verdict as RED. | Display and classification of the repository's own output; it forges no parsed field. |
+| **test gap** | A state-matrix cell without a test, a coverage percentage, a weak or flaky test, a fixture that shares the coordinator's environment. | Category `testing`, below the floor by definition above; a gap is work for the test audit, not a defect in the diff. |
+
+A reviewer that wants a class promoted argues it against the floor with a
+reaching input, in the finding, and the owner rules; the table records the
+rulings, it does not close them.
+
 ## `NO FINDINGS` is a result
 
 A reviewer that finds nothing answers `NO FINDINGS` and nothing else. That is a
@@ -171,7 +196,7 @@ trailers, appended }`.
 
 ## Where this is enforced
 
-- `scripts/resolvers/review-army.ts`: the floor and `NO FINDINGS` in every specialist and Red Team prompt (rendered into `review/sections/review-army.md` and `ship/sections/review-army.md`); in `/ship` every specialist and Red Team prompt it renders carries its diff base as a literal sha the parent writes in (`DIFF_BASE=<diff-base>`: the merge base on pass 1, `RANGE_BASE` from pass 2 on), because a reviewer is its own process and never sees the parent's shell variables; the fan-out batches above.
+- `scripts/resolvers/review-army.ts`: the floor and `NO FINDINGS` in every specialist and Red Team prompt (rendered into `review/sections/review-army.md` and `ship/sections/review-army.md`), and the lead judgment's "On record" line that turns a [ruled class](#rulings-below-the-floor) into a note; in `/ship` every specialist and Red Team prompt it renders carries its diff base as a literal sha the parent writes in (`DIFF_BASE=<diff-base>`: the merge base on pass 1, `RANGE_BASE` from pass 2 on), because a reviewer is its own process and never sees the parent's shell variables; the fan-out batches above.
 - `ship/sections/review-army.md` Step 9: passes, `RANGE_BASE`, classify, defer, render, the cap check before any edit, the Step 11 late-defer row, scratch files under `.gstack/tmp/`, the cap decisions; `ship/sections/adversarial.md` Step 11: the outside voice reads the same range and scopes queued fixes the same way.
 - `lib/review-scope.ts`, `bin/gstack-review-scope`: `classify`, `defer`, `render`, `summary`; refusals are `REVIEW_SCOPE_INPUT` and `REVIEW_SCOPE_RANGE` ([troubleshooting.md](troubleshooting.md#review-scope-input)).
 - `lib/ship-policy.ts`: the `review_scope` key and its `ship: review_scope …` consequence line.

docs/ship-policy.md

@@ -425,7 +425,18 @@ conflicts with the base, CHANGELOG-only conflicts included) is replaced by a
 scratch merge of the head onto the base, pushed as `screen/<pr>-<head12>`
 (the `SCREEN_MERGE:` line says which); a screen dispatched on a stale ref would
 plan no cases and report success. Only a real conflict with the base (or a
-remote whose refs cannot be listed) is `QUEUE_SCREEN_STALE_REF`. A finished screen that planned no cases (no
+remote whose refs cannot be listed) is `QUEUE_SCREEN_STALE_REF`. Under either
+choice `ready` first reads what `gstack-pr-wait`'s base-overlap event reports,
+from git: when the base moved since the head forked and its new commits touch
+files the PR changes, `BASE_OVERLAP: <base> moved to <sha7> and touches files
+in this PR: <files>; the train's merge onto <base> is clean` (or `conflicts
+beyond the CHANGELOG and the stamp files (<files>)`), the train's own merge
+rules run as a dry run in object storage; a git read that fails prints
+`conflicts unknown` and refuses nothing. A conflict under `--no-screen` is
+`QUEUE_TRAIN_CONFLICT` at READY, since the train would stop at this PR: the
+lane reports it to the controller and, on its word (crew template rule 14),
+merges the base into the pin, resolves, pushes and runs /ship before READY,
+instead of learning it when the train is built. A finished screen that planned no cases (no
 `paid-plan` artifact, which a superseded run never uploads, or a plan manifest
 with nothing planned) is never followed as a screen: `ready --screen`
 dispatches a fresh one and `train build` prints `SCREEN: #<n> run <id> planned
@@ -453,7 +464,14 @@ and waits (`--free-timeout <s>`, default 2700) for a green
 through eval-bg's nonce. A train fills at `max_size` READY PRs or when its
 oldest READY has waited `fill_hours`, whichever comes first; until then
 `QUEUE_TRAIN_WAITING` names the time, and `--urgent` dispatches a one-PR
-train at once. gstack's own block fills at 3 PRs or one hour (it was 5 PRs or
+train at once. The scratch merges run before the fill rule: every `train
+build` prints `#<pr> BASE_OVERLAP: <base> moved to <sha7> and touches files in
+this PR: <files>` for a seated pin whose files the base moved onto since it
+forked (the chain's merge onto the pin's own prefix is the verdict), refuses
+`QUEUE_TRAIN_CONFLICT` for a pin that does not merge, and otherwise prints
+`MERGE_CHECK: T1..T<k> merge clean onto <base>@<sha>`, so a
+`QUEUE_TRAIN_WAITING` result is also a clean dry-run merge of the pool (the
+crew template's sibling trial merge, done by the tool). gstack's own block fills at 3 PRs or one hour (it was 5 PRs or
 three hours until 2026-10-26, which left READY PRs waiting most of a day). The same pins on the same base return the live train
 instead of a second one; `train build --train <nonce>` resumes a train from
 its record. Refusals, all before any spend: `QUEUE_TRAIN_UNTRUSTED` (a fork
@@ -708,7 +726,9 @@ integers under `review_loop`, `gate` and `ci`: `passes=2`, `passes: 2` and
 several counters, `passed` under `gate` and `ci` or `pass` under every gate
 lane, matches when the number equals any of them; an underscore in a key
 matches `_`, a space or `-`; a name inside a longer word, `3 pass-through`, is
-not a claim). A body that quotes a different number is
+not a claim; a number that is the value of a `key=` or `key:` pair is that
+key's claim alone, so `passes=2 blocking_found=1` is two claims, never a
+`2 blocking_found` one, #3513). A body that quotes a different number is
 `RECEIPT_PROSE_MISMATCH` (exit 1) and the `PROSE:` line names each field with
 both values; prose that quotes no counter is not a claim. Data PRs (evals,
 measurements, sweeps) run it before READY so the body's numbers are the run's.
@@ -798,7 +818,11 @@ dependency, a script or a new key makes them product). A path the policy's
 `stamp_files`, `stamp_paths` or `release_outputs` name is metadata only while
 its change is a stamp, the rule the tree receipt applies to the same paths: a
 manifest by its version key, any other file when its new text reads as the old
-once the new release version is replaced by the old one. A logic change in a
+once the new release version is replaced by the old one, or when every changed
+line is a stamp-form rewrite to the new version of the old version or of a
+newer tentative one the stamp moved (a `*_KEY_VERSION` written as the next
+minor, #3512; a version older than the old one or of another width is a hand
+edit). A logic change in a
 stamped source file is product. Lockfiles are product unless the policy names
 them. A base merge contributes no product
 change of its own. A receipt carried twice keeps the reviewed head as
@@ -940,7 +964,10 @@ only, where `gstack-version-bump` writes; a dependency entry is `changed`),
 the CHANGELOG (a heading move or re-heading, compared as an order-independent
 multiset of entries), `stamp_paths` and `release_outputs` (every
 removed/added pair must be an exact `(was X)` / `since: X` / `vX` rewrite (a quote around X in the first two forms included) with
-X the gated version and the replacement the current one, so a dependency pin
+X the gated version, or a newer version of the repo's width that the
+stamp_paths stage moves (a tentative `*_KEY_VERSION` written as the next
+minor, #3512; a version older than the gated one, or of another width, is a
+hand edit), and the replacement the current one, so a dependency pin
 that happens to equal the old version is `changed`; `stamp_patterns`
 templates count as stamp forms too). With `stamp_files` set, a file outside it
 is `changed (stamp_files: not in the repo's stamp_files allowlist)` whatever

docs/skills.md

@@ -735,6 +735,17 @@ Every `/ship` run builds a code path map from your diff, searches for correspond
 
 A lot of branches die when the interesting work is done and only the boring release work is left. Humans procrastinate that part. AI should not.
 
+### The PR body
+
+`/ship` writes the PR body from the diff, the receipt and the review, and a reader who was not in the session decides the merge from it. On Capy (`CAPY_PR: on`) the body opens with four sections, in this order, before the standard Summary, Test plan and Documentation sections:
+
+1. **`## What changed`**: one to three captioned before/after images from `gstack-pr-shots` or a UI screenshot, highest-signal first, with a collapsed text version under each.
+2. **`## Why`**: two or three sentences. The first states the problem in the person's words (the issue, the message, the failing run), then the cause and the user-visible effect.
+3. **`## Blast radius`**: what else the change touches and the one fact that makes that safe, with the evidence-ladder rung that fact reached (a test that exercises the shared path, a grep over every caller, a measured run). A change to a shared helper, a generated file, a selection list or a policy default always has one; a leaf fix says so in one line.
+4. **`## Approvals`**: the table from `gstack-approvals render`, verbatim.
+
+Before the body is composed, `gstack-prose-lint` reads the Why, Summary and Blast radius drafts (`--sections 'Why,Summary,Blast radius' --gate`) and the CHANGELOG entry (`--base origin/<base> --gate`), and a `PROSE_LINT_FOUND` line names the sentence to rewrite; the rest of the body is advisory. The Captain's merge-ask template (`captain/templates/merge-ask.md`) carries the same four sections, so a READY report and its PR say the same thing.
+
 ### Versions: where `/ship` looks, and when it ships without one
 
 `/ship` bumps a version only when the project says where the version lives. It checks, in order:

docs/standalone-binaries.md

@@ -18,7 +18,7 @@ Beyond the slash-command skills, gstack ships standalone CLIs for workflows that
 | `gstack-review-log` | **Review-pass receipts** — `--start <skill>` captures the working-tree fingerprint before a diff review; `'<JSON>' --finish <token>` consumes that single-use, repository/branch/skill-scoped receipt. Binding requires matching start/end content and reviewer-reported `completed:true` and `converged:true`; it is not independent proof that a model read the source. `--check-shared-libs <token>` reads a current finding as JSON on stdin and checks prior Skip decisions against the actual branch, capture and eligible raw source blobs. It returns `reusable:false` when proof is missing or unsafe, including older records without logger-versioned coverage. Final review logging computes shared-code fingerprints and coverage rather than trusting supplied proof. |
 | `gstack-review-read` | **Review freshness** — emits review records with computed `review_freshness.status` and `reason`: CURRENT, STALE, or UNVERIFIED for diff reviews. `/ship` and `/land-and-deploy` use the same grade; a matching commit alone never certifies a diff review. [Dashboard rules](skills.md#review-readiness-dashboard). |
 | `gstack-evidence` | **Verification-evidence ledger** — `run --label <lane> -- <cmd>` transparently wraps any test command (the child's exit code always passes through) and records what ran against which working-tree fingerprint; `check` grades each label FRESH/STALE/MISSING with `--expect-cmd`, `--max-age`, and `--allow-paths` binding. /ship and /land-and-deploy cite fresh evidence instead of re-running suites. Per-run logs are 0600, capped at 2MB, pruned after 30 days; the ledger and logs stay machine-local by design. |
-| `gstack-pr-wait` | **PR event watcher** — `gstack-pr-wait <pr> [--repo owner/name] [--until green\|ready\|any-change] [--interval 20] [--timeout 60]` polls one pull request through `gh` and prints one stdout line per real event (a check failed with its link, all checks finished, a comment or review from someone other than the author, a review bot finished on the head, the base moved and its new commits touch this PR's files, BEHIND or DIRTY, merged or closed), with progress on stderr, so a monitor wakes the agent once per event instead of on a timer. `green` gates on every check on the head (an advisory bot's review is never waited for, its checks gate like any other), as `gstack-ci-gate` does; a check the base branch requires (rulesets and classic branch protection) is pending until it appears, never green by omission; two identical polls confirm a check set is complete. Exit 0 when the condition is met, 1 when it needs action or `--timeout` passes (`PR_WAIT_BLOCKED`), 2 on bad input, a second watcher for the same PR or a local error (`PR_WAIT_REFUSED`). State under `<state root>/pr-wait/`, baseline discarded after 6 hours. Review bots are configuration: `gstack-config set pr_wait_gating_bots greptile,macroscope`, `pr_wait_advisory_bots coderabbit` (advisory bots are reported, never waited on). `/land-and-deploy` Step 3 is one call of it. Design ported from tstack's pr-wait (MIT, Theo Browne; NOTICE.md). [Exit codes](troubleshooting.md#gstack-pr-wait). |
+| `gstack-pr-wait` | **PR event watcher** — `gstack-pr-wait <pr> [--repo owner/name] [--until green\|ready\|any-change] [--interval 20] [--timeout 60] [--exclude CHECK,...]` polls one pull request through `gh` and prints one stdout line per real event (a check failed with its link, all checks finished, a comment or review from someone other than the author, a review bot finished on the head, the base moved and its new commits touch this PR's files, BEHIND or DIRTY, merged or closed), with progress on stderr, so a monitor wakes the agent once per event instead of on a timer. `green` gates on every check on the head (an advisory bot's review is never waited for, its checks gate like any other), as `gstack-ci-gate` does; a check the base branch requires (rulesets and classic branch protection) is pending until it appears, never green by omission; two identical polls confirm a check set is complete; `--exclude` leaves a named non-required check out of the verdict as `gstack-ci-gate --exclude` does (a required name is refused and gates). Exit 0 when the condition is met, 1 when it needs action or `--timeout` passes (`PR_WAIT_BLOCKED`), 2 on bad input, a second watcher for the same PR or a local error (`PR_WAIT_REFUSED`). State under `<state root>/pr-wait/`, baseline discarded after 6 hours. Review bots are configuration: `gstack-config set pr_wait_gating_bots greptile,macroscope`, `pr_wait_advisory_bots coderabbit` (advisory bots are reported, never waited on). `/land-and-deploy` Step 3 is one call of it. Design ported from tstack's pr-wait (MIT, Theo Browne; NOTICE.md). [Exit codes](troubleshooting.md#gstack-pr-wait). |
 | `gstack-issue-guard` | **Tracker-text trust envelope** — fetches GitHub issue/PR text (`issue <n>`, `pr-body`, `pr-comments`, or `--stdin`) and wraps it in a labeled envelope so agents treat it as data: injection-shaped lines get labeled even through fullwidth and invisible-character evasion, and forged envelope banners are defused. Every tracker-text ingress in gstack routes through it, enforced by a CI scanner. |
 | `gstack-ios-qa-daemon` | **iOS QA daemon** — Mac-side broker between an agent and a connected iPhone or iPad over USB CoreDevice. Loopback by default; `--tailnet` opens a Tailscale-facing listener with identity-gated capability tiers. Single-instance via flock on `~/.gstack/ios-qa-daemon.pid`. See [docs/howto-ios-testing-with-gstack.md](howto-ios-testing-with-gstack.md). |
 | `gstack-ios-qa-mint` | **iOS allowlist manager** — owner-grant CLI for the tailnet allowlist. `grant`/`revoke`/`list` against `~/.gstack/ios-qa-allowlist.json` (mode 0600). Remote agents never auto-allowlist; this is the explicit-intent path. |

docs/troubleshooting.md

@@ -2373,7 +2373,12 @@ exact tokens it finds in these positions:
   `gates`, `backstop`, `bisect`, `rerun`, `dispatched`, `dispatch`,
   `failed`, `passed` (`failed 40269992077 | rerun passed 40287239206`,
   `gate 40323989742`, `run #19368712345`); a generic word such as `id`,
-  `job`, `attempt` or `workflow` exempts nothing;
+  `job`, `attempt` or `workflow` exempts nothing. This position is exempt in
+  the scanner itself (`isCiRunIdentifier` in `lib/redact-patterns.ts`), so
+  `gstack-approvals add`, `gstack-decision-trail` and the pregate secrets
+  check accept it too, not only `gstack-post`. A national number written
+  without its `+` right after a run word reads as a run id there; that is the
+  accepted residual of the ruling that a run id is not a phone number;
 - every 11-digit id chained after an allowlisted one through `, `, ` and `,
   ` & `, ` or ` or ` / `, optionally with a `(#1234)` between
   (`screen 40269992077 (#3434) and 40275443393 (#3453)`).
@@ -2708,11 +2713,19 @@ code (you already do when you run its tests), or drop the release tool.
 **Meaning.** `gstack-restamp` was told the base SHA (`--expect-base`) or the
 merge order (`--expect-order`) the allocation assumed, and one of them moved:
 the base advanced, a predecessor is not merged, or a successor merged first.
-The JSON result carries `expected`, `actual` and `recompute`.
+`--expect-order` lists the lane's own PR (`--pr`) first, then the remaining
+order; a list that leaves it out reads every entry as an unmerged predecessor,
+and the message says so. The JSON result carries `expected`, `actual` and
+`recompute`.
 
-**Fix.** Run the printed `recompute` command, which recomputes the version
-from the current base, then merge with the server-side head condition
-(`MERGE_CONDITION: sha=<head>`).
+**Fix.** Run the printed `recompute` command: it carries the corrected order
+(this PR first, merged PRs dropped, an unmerged predecessor kept ahead because
+it must merge first) and recomputes the version from the current base; then
+merge with the server-side head condition (`MERGE_CONDITION: sha=<head>`). A
+predecessor GitHub calls merged that the local base does not carry yet
+(`#<pr>=merged-not-in-base`) is kept ahead too, with `fix: git fetch origin
+<base>`: fetch, then rerun, so the recompute, pinned to that local base, never
+allocates the version the predecessor already took.
 
 <a id="restamp-version-source"></a>
 ### `RESTAMP_VERSION_SOURCE`
@@ -4174,13 +4187,19 @@ with the new pin (`train build --prs <n>@<new sha>,…`).
 ### `QUEUE_TRAIN_CONFLICT`
 
 **Meaning.** Merging a PR onto the prefix before it conflicts outside the
-version files and the CHANGELOG. The `QUEUE_TRAIN_CONFLICT:` line names the
-PR, the conflicting files and the PRs merged before it. Nothing was
-dispatched and no branch was pushed.
+version files and the CHANGELOG. From `train build`, the
+`QUEUE_TRAIN_CONFLICT:` line names the PR, the conflicting files and the PRs
+merged before it, and the check runs before the fill rule, so a filling train
+finds it too. From `ready --no-screen`, the prefix is the base itself: the
+`BASE_OVERLAP:` line names the files the base moved onto since the head
+forked, and the train would stop at this PR (under `--screen` the same
+conflict is `QUEUE_SCREEN_STALE_REF`). Nothing was dispatched and no branch
+was pushed.
 
-**Fix.** Run the printed `train build --prs <list without that PR>`. The PR's
-lane merges main (with the earlier PRs once they land) and reports READY
-again.
+**Fix.** From `train build`, run the printed `train build --prs <list without
+that PR>`. The PR's lane reports the conflict to the controller and, on its
+word (crew template rule 14), merges main into the pin (with the earlier PRs
+once they land), resolves, pushes, runs /ship and reports READY again.
 
 <a id="queue-train-free-red"></a>
 ### `QUEUE_TRAIN_FREE_RED`
@@ -4209,7 +4228,11 @@ run carrying the nonce, or marks the dispatch absent; only then does
 
 **Meaning.** The train has fewer PRs than `train.max_size` and its oldest
 READY PR has waited less than `train.fill_hours`, so it does not dispatch
-yet. The message prints when it will.
+yet. The message prints when it will. The scratch merges already ran: the
+`MERGE_CHECK: T1..T<k> merge clean onto <base>@<sha>` line above it says every
+seated pin merges in slot order (a pin that does not is `QUEUE_TRAIN_CONFLICT`
+instead), and a `#<pr> BASE_OVERLAP:` line names a pin whose files the base
+moved onto since it forked.
 
 **Fix.** Add READY PRs, rerun after the printed time, or pass `--urgent` for a
 one-PR train.
@@ -4688,14 +4711,21 @@ reason }` and carry the `Documentation exception: accepted by <owner> <date> via
 ## gstack-pr-wait
 
 `gstack-pr-wait <pr> [--repo owner/name] [--until green|ready|any-change]
-[--interval SECONDS] [--timeout MINUTES]` watches one pull request and prints
+[--interval SECONDS] [--timeout MINUTES] [--exclude CHECK,...]` watches one pull request and prints
 one stdout line per real event, prefixed `#<pr>`; progress and recoverable
 errors go to stderr. `/land-and-deploy` Step 3 runs it with `--until green`.
 Exit 0 means the `--until` condition is met or the PR merged; exit 1 and 2
 end with one of the two codes below. `green` gates on every check on the
 head, as `gstack-ci-gate` does, so the two never disagree; a check the base
 branch requires (rulesets and classic branch protection) counts as pending
-until it appears. A 403 on the classic
+until it appears. `--exclude <check,...>` leaves a named non-required check
+out of the verdict, as `gstack-ci-gate --exclude` does: it neither holds
+`green` while pending nor blocks when it fails, and the first poll logs
+`excluded from the verdict: <names>`; a required check cannot be excluded
+(`exclude refused for required check <name>`) and gates, and while a
+required-check source cannot be read nothing is excluded (`exclude refused for
+<names>: the required checks could not be read`), as `gstack-ci-gate` refuses
+its overrides. A 403 on the classic
 protection read (a token without admin scope) is logged and skipped, and the
 rulesets read still names the required checks. The review-bot lists are
 configuration: `gstack-config set pr_wait_gating_bots greptile,macroscope`

lib/gstack-post.ts

@@ -21,6 +21,7 @@
  */
 import { createHash } from "crypto";
 import { scan, type Finding, type RepoVisibility } from "./redact-engine";
+import { CI_RUN_WORD_BEFORE } from "./redact-patterns";
 import { RECEIPT_FENCE } from "./ship-receipt";
 import { canonicalRemote } from "./remote-identity";
 import { parseChangelog } from "./changelog-check";
@@ -329,8 +330,6 @@ function repoReleaseVersions(env: PostEnv): string[] {
   return [...versions];
 }
 
-/** A run word with an optional ':' '#' '=' right before an 11-digit token: `run 19368712345`, `gate=40323989742`, `rerun passed 40287239206`. */
-const RUN_WORD_BEFORE = /\b(?:runs?|screens?|gates?|backstop|bisect|rerun|dispatch(?:ed)?|failed|passed)\s*[:#=]?\s*$/i;
 /** Between two 11-digit tokens of one list: `, `, ` and `, ` & `, ` or `, ` / `, with an optional `(#1234)` after the first. */
 const RUN_ID_CHAIN = /^(?:\s*\(#\d+\))?\s*(?:,|and|&|or|\/)\s*$/i;
 
@@ -344,7 +343,9 @@ const RUN_ID_CHAIN = /^(?:\s*\(#\d+\))?\s*(?:,|and|&|or|\/)\s*$/i;
  * word on the same line (run, runs, screen, screens, gate, gates, backstop,
  * bisect, rerun, dispatched, dispatch, failed, passed, with an optional ':'
  * '#' '=' between: `failed 40269992077`, `gate 40323989742`,
- * `run #19368712345`; generic words such as id, job, attempt and workflow
+ * `run #19368712345`; the engine's isCiRunIdentifier exempts that position on
+ * its own since the run-id fix for approval records, and the same regex drives
+ * the chain below; generic words such as id, job, attempt and workflow
  * exempt nothing, since `customer id 14155551234` is a number), plus every 11-digit id chained
  * after an allowlisted one through ", ", " and ", " & ", " or " or " / ",
  * optionally with a "(#1234)" between (`screen 40269992077 (#3434) and
@@ -361,7 +362,7 @@ function receiptDigitTokens(texts: string[]): string[] {
       let chainFrom = -1;
       for (const m of line.matchAll(/(?<![\d+])\d{11}(?!\d)/g)) {
         const at = m.index!;
-        const listed = RUN_WORD_BEFORE.test(line.slice(0, at)) || (chainFrom >= 0 && RUN_ID_CHAIN.test(line.slice(chainFrom, at)));
+        const listed = CI_RUN_WORD_BEFORE.test(line.slice(0, at)) || (chainFrom >= 0 && RUN_ID_CHAIN.test(line.slice(chainFrom, at)));
         if (listed) tokens.add(m[0]);
         chainFrom = listed ? at + m[0].length : -1;
       }

lib/pr-wait-github.ts

@@ -193,24 +193,28 @@ export function fetchChecks(env: GhEnv, repo: string, pr: number): Map<string, C
  * Required check names for the base branch, from rulesets and classic branch
  * protection. Reading them up front means a required check that has not
  * started yet counts as pending instead of green by omission. A 404 means no
- * protection of that kind, which is normal; any other failure is logged and
- * that source contributes nothing.
+ * protection of that kind, which is normal; any other failure is logged, that
+ * source contributes nothing, and `known` is false: the set may be short, so a
+ * caller that would relax the gate on it (`--exclude`) must not.
  */
-export function requiredChecks(env: GhEnv, repo: string, base: string): Set<string> {
+export function requiredChecks(env: GhEnv, repo: string, base: string): { names: Set<string>; known: boolean } {
   const sources: Array<[string, string]> = [
     [`repos/${repo}/rules/branches/${base}`, '.[] | select(.type == "required_status_checks") | .parameters.required_status_checks[].context'],
     [`repos/${repo}/branches/${base}/protection/required_status_checks`, '.contexts[]?, .checks[]?.context'],
   ];
   const names = new Set<string>();
+  let known = true;
   for (const [path, jq] of sources) {
     try {
       for (const name of ghLines(env, ['api', path, '--jq', jq])) names.add(name);
     } catch (error) {
       const message = error instanceof Error ? error.message : String(error);
-      if (!/HTTP 404/.test(message)) env.log(`could not read required checks from ${path}: ${message}`);
+      if (/HTTP 404/.test(message)) continue;
+      known = false;
+      env.log(`could not read required checks from ${path}: ${message}`);
     }
   }
-  return names;
+  return { names, known };
 }
 
 /** Files the PR changes, from `gh pr diff --name-only`. */

lib/pr-wait.ts

@@ -46,12 +46,22 @@ export interface PrWaitEnv {
   pidAlive(pid: number): boolean;
 }
 
-export interface PrWaitOptions { target?: string; repo?: string; until: Until; interval: number; timeout: number }
+export interface PrWaitOptions { target?: string; repo?: string; until: Until; interval: number; timeout: number; excludes: string[] }
 export interface BotLists { gating: string[]; advisory: string[] }
 
 interface State { reported: string[]; summaries: Record<string, string>; base_oid: string | null; merge_state: string | null }
 
-export const PR_WAIT_USAGE = `gstack-pr-wait <pr> [--repo owner/name] [--until green|ready|any-change] [--interval SECONDS] [--timeout MINUTES]
+/**
+ * The base-overlap event, one shape for the watcher and for `gstack-ship-queue
+ * ready` / `train build`, which read the same fact from git (handoff
+ * GSTA-119): the base moved and its new commits touch files this PR changes.
+ */
+export function baseOverlapLine(base: string, baseOid: string, touched: readonly string[]): string {
+  const more = touched.length > 5 ? ` and ${touched.length - 5} more` : '';
+  return `${base} moved to ${baseOid.slice(0, 7)} and touches files in this PR: ${touched.slice(0, 5).join(', ')}${more}`;
+}
+
+export const PR_WAIT_USAGE = `gstack-pr-wait <pr> [--repo owner/name] [--until green|ready|any-change] [--interval SECONDS] [--timeout MINUTES] [--exclude CHECK,...]
 
 Watches one pull request and prints one line per real event on stdout (prefixed
 #<pr>), so a monitor wakes the agent once per event. Progress and recoverable
@@ -69,6 +79,10 @@ comes from the URL or the current checkout.
 
 --timeout   minutes before a watch that met nothing ends as PR_WAIT_BLOCKED,
             naming what is still pending (default 60; 0 waits without limit)
+--exclude   check names (comma-separated, repeatable) left out of the verdict,
+            as gstack-ci-gate --exclude does: a named check neither holds green
+            while pending nor blocks when it fails. A check the base branch
+            requires cannot be excluded; the name is logged as refused and gates
 
 Events: a check failed (name and link); all checks finished on the
 head (one summary line); a comment, review, thread or thread reply from someone
@@ -101,7 +115,7 @@ lock, which the next watcher for the PR takes over once the pid is gone.
 Docs: docs/troubleshooting.md#gstack-pr-wait`;
 
 export function parsePrWaitArgs(argv: string[]): PrWaitOptions | { help: true } | { error: string } {
-  const opts: PrWaitOptions = { until: 'ready', interval: 20, timeout: DEFAULT_TIMEOUT_MINUTES };
+  const opts: PrWaitOptions = { until: 'ready', interval: 20, timeout: DEFAULT_TIMEOUT_MINUTES, excludes: [] };
   for (let i = 0; i < argv.length; i++) {
     const flag = argv[i];
     if (flag === '--help' || flag === '-h') return { help: true };
@@ -124,6 +138,12 @@ export function parsePrWaitArgs(argv: string[]): PrWaitOptions | { help: true }
       opts.timeout = Number(value);
       continue;
     }
+    if (flag === '--exclude') {
+      const names = (argv[++i] ?? '').split(',').map(s => s.trim()).filter(Boolean);
+      if (!names.length) return { error: '--exclude needs one or more check names, comma-separated' };
+      for (const name of names) if (!opts.excludes.includes(name)) opts.excludes.push(name);
+      continue;
+    }
     if (flag.startsWith('-') && !/^-?\d+$/.test(flag)) return { error: `unknown argument ${flag}` };
     if (opts.target !== undefined) return { error: `unexpected argument ${flag}` };
     opts.target = flag;
@@ -185,12 +205,14 @@ export class Watcher {
   /** First run for this PR: record what exists and report nothing. */
   baseline: boolean;
   private required: Set<string> | null = null;
+  /** False when a required-check source could not be read: the set may be short, so no --exclude name is excluded. */
+  private requiredKnown = true;
   private prevSnapshot: string | null = null;
   private prFiles = new Map<string, Set<string>>();
   private pending: string[] = [];
   private readonly gh: GhEnv;
 
-  constructor(readonly env: PrWaitEnv, readonly repo: string, readonly pr: number, readonly until: Until, readonly interval: number, readonly bots: BotLists, readonly timeoutMinutes = DEFAULT_TIMEOUT_MINUTES) {
+  constructor(readonly env: PrWaitEnv, readonly repo: string, readonly pr: number, readonly until: Until, readonly interval: number, readonly bots: BotLists, readonly timeoutMinutes = DEFAULT_TIMEOUT_MINUTES, readonly excludes: readonly string[] = []) {
     const stem = path.join(stateDir(env.vars), `${repo.replace('/', '__')}__${pr}`);
     this.stateFile = `${stem}.json`;
     this.lockFile = `${stem}.lock`;
@@ -295,11 +317,22 @@ export class Watcher {
     return count === null ? '' : `; ${count} commit${count === '1' ? '' : 's'} behind ${base}`;
   }
 
-  private evaluateChecks(checks: Map<string, Check>, head: string) {
+  /**
+   * `--exclude` names that are not required checks: the ones the verdict leaves out, as gstack-ci-gate does. While the
+   * required set is unknown (a source could not be read) nothing is excluded, as gstack-ci-gate refuses its overrides.
+   */
+  private excluded(): Set<string> {
+    return new Set(this.requiredKnown ? this.excludes.filter(name => !this.required?.has(name)) : []);
+  }
+
+  private evaluateChecks(all: Map<string, Check>, head: string) {
     const required = this.required ?? new Set<string>();
     // Every check on the head gates, as gstack-ci-gate does, so land-and-deploy's two gates never
     // disagree; a required check gates even before it appears. An advisory bot's review is never
-    // waited for, but its check is a check like any other.
+    // waited for, but its check is a check like any other. A --exclude name that is not required
+    // is left out of the verdict; a required one cannot be excluded and gates.
+    const excluded = this.excluded();
+    const checks = new Map([...all].filter(([name]) => !excluded.has(name)));
     const watched = checks;
     const gate = new Set([...required, ...watched.keys()]);
     const bucketOf = (name: string) => checks.get(name)?.bucket ?? 'pending';
@@ -392,10 +425,7 @@ export class Watcher {
     const oldBase = this.state.base_oid;
     if (pr.baseOid && oldBase && pr.baseOid !== oldBase && !this.baseline) {
       const touched = this.baseOverlap(head, oldBase, pr.baseOid);
-      if (touched.length > 0) {
-        const more = touched.length > 5 ? ` and ${touched.length - 5} more` : '';
-        this.emit(`base:${pr.baseOid}`, `${base} moved to ${pr.baseOid.slice(0, 7)} and touches files in this PR: ${touched.slice(0, 5).join(', ')}${more}`);
-      }
+      if (touched.length > 0) this.emit(`base:${pr.baseOid}`, baseOverlapLine(base, pr.baseOid, touched));
     }
     if (pr.baseOid) this.state.base_oid = pr.baseOid;
     const mergeState = pr.mergeStateStatus;
@@ -418,9 +448,19 @@ export class Watcher {
     if (pr.state === 'CLOSED') return this.finish(PR_WAIT_EXIT.blocked, [`closed without merging: ${pr.url}`]);
     const base = pr.baseRefName;
     if (this.required === null) {
-      this.required = requiredChecks(this.gh, this.repo, base);
+      const read = requiredChecks(this.gh, this.repo, base);
+      this.required = read.names;
+      this.requiredKnown = read.known;
       const gate = [...this.required].sort().join(', ') || 'none set';
       this.log(`watching ${this.repo}#${this.pr} until ${this.until}; required checks: ${gate}; every check on the head gates`);
+      if (this.excludes.length && !this.requiredKnown) {
+        this.log(`exclude refused for ${this.excludes.join(', ')}: the required checks could not be read, so every check gates`);
+      } else if (this.excludes.length) {
+        const refused = this.excludes.filter(name => this.required!.has(name));
+        const dropped = this.excludes.filter(name => !this.required!.has(name));
+        if (dropped.length) this.log(`excluded from the verdict: ${dropped.join(', ')}`);
+        if (refused.length) this.log(`exclude refused for required check${refused.length === 1 ? '' : 's'} ${refused.join(', ')}: a required check gates`);
+      }
     }
     const checks = fetchChecks(this.gh, this.repo, this.pr);
     const head = pr.headRefOid;
@@ -530,7 +570,7 @@ export function prWaitMain(argv: string[], env: PrWaitEnv): number {
   if ('error' in parsed) return refuse(env, parsed.error);
   const target = resolveTarget(env, parsed);
   if ('error' in target) return refuse(env, target.error);
-  const watcher = new Watcher(env, target.repo, target.pr, parsed.until, parsed.interval, botLists(env.vars), parsed.timeout);
+  const watcher = new Watcher(env, target.repo, target.pr, parsed.until, parsed.interval, botLists(env.vars), parsed.timeout, parsed.excludes);
   // Exit 1 always carries a PR_WAIT_BLOCKED line, so a local fault (an unwritable state root, a full disk) is a refusal, never a verdict.
   try {
     const lock = watcher.lock();

lib/pregate/import-graph.ts

@@ -49,7 +49,11 @@ const scanner = new Bun.Transpiler({ loader: 'tsx' });
 // `( ) [ ] +` appear in framework route paths: Next.js route groups `(app)`,
 // dynamic segments `[id]` / `[[...slug]]`, SvelteKit `+page.svelte`. A literal
 // that matches more text only gains an edge when it resolves to a repo file.
-const QUOTED = /['"`]([A-Za-z0-9_@][A-Za-z0-9_./@()[\]+-]*)['"`]/g;
+// A relative literal (`'../ship/sections/pr-body.md.tmpl'` read through
+// `new URL(…, import.meta.url)` or `readFileSync`) resolves from the file that
+// names it, like a relative import; without it a test reading a template by
+// relative path had no edge to the template (GSTA-121's test/ship-publication-gates.test.ts).
+const QUOTED = /['"`]((?:\.\.?\/)*[A-Za-z0-9_@][A-Za-z0-9_./@()[\]+-]*)['"`]/g;
 const JOIN = /\bjoin\(([^()]*)\)/g;
 const RESOLVE_EXT = ['', '.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs', '.mts', '.cts', '/index.ts', '/index.tsx', '/index.js', '/index.jsx', '/index.mjs'];
 // A statement starts a line or follows `;` or a compound statement's `:`. The module spec is

lib/redact-patterns.ts

@@ -328,16 +328,28 @@ export function isNumericMetadataValue(span: string, match: RegExpExecArray): bo
  * command. Only the position decides: a run/job flag (`--run 37235771700`,
  * `--job=111534615007`), a `gh run view|watch|rerun|download|cancel`
  * argument, an `actions/runs/<id>` or `/job/<id>` URL segment, or a
- * `run_id` / `job_id` key. The same digits anywhere else still report.
+ * `run_id` / `job_id` key. An 11-digit id (every GitHub Actions run id since
+ * 2025) also counts right after a run word on the same line, the way a ship
+ * receipt, a READY block or an approval record names one (`screen 40477638211
+ * GREEN`, `gate=40323989742`, `rerun passed 40287239206`, `run #19368712345`);
+ * generic words (id, job, attempt, workflow) exempt nothing, since `customer id
+ * 14155551234` is a number. A national number written without its `+` right
+ * after a run word (`failed 14155551234`) is the accepted residual of the
+ * owner's ruling that a run id is not a phone number; gstack-post's chained-id
+ * allowlist reads the same position the same way. The same digits anywhere
+ * else still report.
  */
 const CI_ID_POSITION_BEFORE =
   /(?:--(?:run|job)(?:-id)?(?:[ \t]+|[ \t]*=[ \t]*)|\bgh[ \t]+run[ \t]+(?:view|watch|rerun|download|cancel)[ \t]+|\/actions\/runs\/|\/jobs?\/|(?:run|job)_id["']?[ \t]*[:=][ \t]*["']?)$/i;
+/** A run word with an optional ':' '#' '=' right before an 11-digit token; shared with gstack-post's chained-id allowlist. */
+export const CI_RUN_WORD_BEFORE = /\b(?:runs?|screens?|gates?|backstop|bisect|rerun|dispatch(?:ed)?|failed|passed)\s*[:#=]?\s*$/i;
 export function isCiRunIdentifier(span: string, match: RegExpExecArray): boolean {
   if (!/^\d+$/.test(span)) return false;
   const input = match.input ?? "";
   const { start } = spanBounds(match);
   const lineStart = input.lastIndexOf("\n", start - 1) + 1;
-  return CI_ID_POSITION_BEFORE.test(input.slice(Math.max(lineStart, start - 80), start));
+  const before = input.slice(Math.max(lineStart, start - 80), start);
+  return CI_ID_POSITION_BEFORE.test(before) || (span.length === 11 && CI_RUN_WORD_BEFORE.test(before));
 }
 
 /**

lib/restamp.ts

@@ -128,12 +128,33 @@ function queueOrderCheck(o: RestampOptions, baseSha: string): void {
     actual[pr] = merged ? (inBase ? 'merged' : 'merged-not-in-base') : String(info.state ?? 'unknown').toLowerCase();
   }
   const stale = [...ahead.filter(pr => actual[pr] !== 'merged'), ...behind.filter(pr => actual[pr]?.startsWith('merged'))];
-  if (stale.length) {
-    throw new Refusal('QUEUE_STALE', `queue order moved: ${stale.map(pr => `#${pr}=${actual[pr]}`).join(', ')}`, {
-      expected: { order: o.expectOrder, base: o.expectBase ?? null }, actual: { prs: actual, base: baseSha },
-      recompute: `gstack-restamp --next --base ${o.base} --expect-base ${baseSha}${mine ? ` --pr ${mine}` : ''} --expect-order ${o.expectOrder.join(',')}`,
-    });
-  }
+  if (!stale.length) return;
+  // The recompute command is the corrected order, never the stale list again: this PR first (a list that left it out read
+  // every entry as a predecessor), PRs merged into the local base dropped, an unmerged predecessor kept ahead because it
+  // still has to merge first. A predecessor GitHub calls merged that the local base does not carry stays ahead too: the
+  // recompute pins --expect-base to that local base, and dropping the PR would let it allocate the version the
+  // predecessor already took (pre-landing review, pass 1); the lane fetches the base and the entry reads merged.
+  const corrected = correctedOrder(o.expectOrder, mine, actual);
+  const waitingOn = ahead.filter(pr => actual[pr] !== 'merged' && actual[pr] !== 'merged-not-in-base');
+  const unfetched = ahead.filter(pr => actual[pr] === 'merged-not-in-base');
+  const why = mine && idx < 0 ? `; --pr ${mine} is not in --expect-order, so every listed PR read as a predecessor: list this PR first, then the remaining order`
+    : [waitingOn.length ? `${waitingOn.map(pr => `#${pr}`).join(', ')} must merge before this PR stamps` : '',
+      unfetched.length ? `${unfetched.map(pr => `#${pr}`).join(', ')} merged but the local ${o.base} does not carry the merge yet: git fetch origin ${o.base}, then rerun` : '']
+      .filter(Boolean).map(s => `; ${s}`).join('');
+  throw new Refusal('QUEUE_STALE', `queue order moved: ${stale.map(pr => `#${pr}=${actual[pr]}`).join(', ')}${why}`, {
+    expected: { order: o.expectOrder, base: o.expectBase ?? null }, actual: { prs: actual, base: baseSha },
+    recompute: `gstack-restamp --next --base ${o.base} --expect-base ${baseSha.slice(0, 12)}${mine ? ` --pr ${mine}` : ''} --expect-order ${corrected.join(',')}`,
+    ...(unfetched.length ? { fix: `git fetch origin ${o.base}` } : {}),
+  });
+}
+
+/**
+ * The --expect-order to run next: this PR first when the list left it out, every PR merged into the local base dropped,
+ * the rest (open PRs and merges the local base does not carry yet) in their order.
+ */
+export function correctedOrder(order: readonly string[], mine: string | null, actual: Record<string, string>): string[] {
+  const rest = order.filter(pr => actual[pr] !== 'merged');
+  return mine && !order.includes(mine) ? [mine, ...rest] : rest;
 }
 
 interface Plan {
@@ -148,7 +169,7 @@ function planVersions(o: RestampOptions, loaded: LoadedPolicy, policyLabel: stri
   if (o.expectBase && !baseSha.out.startsWith(o.expectBase)) {
     throw new Refusal('QUEUE_STALE', `base moved: expected ${o.expectBase}, ${baseRef} is ${baseSha.out}`, {
       expected: { base: o.expectBase }, actual: { base: baseSha.out },
-      recompute: `gstack-restamp --next --base ${o.base} --expect-base ${baseSha.out.slice(0, 12)}${o.expectOrder?.length ? ` --expect-order ${o.expectOrder.join(',')}` : ''}`,
+      recompute: `gstack-restamp --next --base ${o.base} --expect-base ${baseSha.out.slice(0, 12)}${o.pr ? ` --pr ${o.pr}` : ''}${o.expectOrder?.length ? ` --expect-order ${correctedOrder(o.expectOrder, o.pr ?? null, {}).join(',')}` : ''}`,
     });
   }
   queueOrderCheck(o, baseSha.out);
@@ -313,6 +334,40 @@ export function tentativeMentions(added: Map<string, Set<string>>, forkVersion:
   return [...found];
 }
 
+/** The versions one line mentions in a stamp form (the built-in forms and the policy's `stamp_patterns` templates). */
+export function stampMentions(text: string, templates: readonly string[] = []): Set<string> {
+  const found = new Set<string>();
+  for (const re of [...BUILT_IN_MENTIONS, ...templates.map(templateMention)]) for (const m of text.matchAll(re)) found.add(m.groups!.v!);
+  return found;
+}
+
+/**
+ * A removed/added line pair that a restamp alone explains: every version the
+ * removed line mentions in a stamp form and the added line no longer does
+ * moves to `newV`, and the rewritten line is the added line byte for byte.
+ * The moved versions are the ones the stamp_paths stage moves, not only the
+ * old version `oldV` (the gated head's): a tentative `*_KEY_VERSION = '<next
+ * minor>'` never equals the base version at the slot and is a stamp all the
+ * same (#3512). The stage's own bound applies: a tentative mention is newer
+ * than the version the branch was written against, so a moved version older
+ * than `oldV` (a released key moved by hand) is not a stamp, and neither is
+ * one of another width (`tentativeMentions`: a 3-digit dependency pin in a
+ * 4-digit repo is never a candidate, so moving it is a hand edit). A line that
+ * moves nothing, or changes anything beyond those mentions, is not one either.
+ */
+export function stampOnlyLine(removed: string, added: string, oldV: string, newV: string, templates: readonly string[] = []): boolean {
+  if (removed === added) return false;
+  const floor = parseVersion(oldV);
+  const width = versionWidth(oldV);
+  const after = stampMentions(added, templates);
+  const moved = [...stampMentions(removed, templates)].filter(v => v !== newV && !after.has(v));
+  if (!moved.length || !floor) return false;
+  if (moved.some(v => { const p = parseVersion(v); return !p || versionWidth(v) !== width || cmpVersion(p, floor) < 0; })) return false;
+  let text = removed;
+  for (const v of moved) text = rewriteStamps(text, v, newV, templates).text;
+  return text === added;
+}
+
 /** The version at the branch's original fork point: the first parent of its oldest own commit (base when it has none). */
 export function forkVersion(repoRoot: string, baseRef: string, fallback: string): string {
   const first = git(repoRoot, ['rev-list', '--reverse', '--first-parent', `${baseRef}..HEAD`]).out.split('\n')[0]?.trim();

lib/ship-queue-train-ready.ts

@@ -28,7 +28,7 @@ import { dispatchCi } from './ship-queue-dispatch-ci';
 import { recordVerdict, screenVerdict, verdictLines, type ScreenOutcome } from './ship-queue-screen';
 import { execFor, repoSlug } from './ship-queue-train-github';
 import { loadReady, nowIso, persist, readyPath, rebuildFromNotes, type ReadyReceipt, type ScreenChoice } from './ship-queue-train-record';
-import { fetchBase, mergeCommit, outOk, outPending, outRefused, trainEnv, type TrainEnv, type TrainOut } from './ship-queue-train';
+import { baseOverlap, fetchBase, mergeCommit, outOk, outPending, outRefused, trainEnv, type TrainEnv, type TrainOut } from './ship-queue-train';
 
 const Q = 'gstack-ship-queue';
 
@@ -181,6 +181,18 @@ export async function trainReady(ctx: QueueContext, o: ReadyOptions): Promise<Tr
   if (kind.kind === 'invalid') return outRefused({ code: 'QUEUE_NOT_READY', message: `the Version Gate refuses ${head.slice(0, 12)}: ${kind.reason}` }, base.lines, `fix the head, push, then ${Q} ready --pr ${o.pr} --screen|--no-screen`, 'none');
   const sel = await screenSelection(env, head, baseSha);
   const lines = [...base.lines, ...sel.lines];
+  // gstack-pr-wait's base-overlap event, read from git at READY (handoff GSTA-119): the lane learns here that the base
+  // moved onto its files, and a merge the train would stop at is refused before the pin is pooled. Under --screen the
+  // scratch merge below reports the same conflict as QUEUE_SCREEN_STALE_REF. A comparison git could not make is printed as
+  // unknown and refuses nothing: the screen or the train makes the merge for real.
+  const overlap = baseOverlap(ctx.repoRoot, ctx.base, baseSha, head, env.policy.policy);
+  if (overlap.line) lines.push(overlap.line);
+  if (overlap.conflicts && o.screen !== 'screen') {
+    // The fix follows the crew template (captain/templates/CREW.md rule 14): a lane merges the base into a READY PR only on
+    // the controller's word, so the refusal is reported and the merge waits for that word.
+    return outRefused({ code: 'QUEUE_TRAIN_CONFLICT', message: `#${o.pr} conflicts with ${ctx.base}@${baseSha.slice(0, 12)} beyond the CHANGELOG and the stamp files (${overlap.conflicts.join(', ')}); the train would stop at this PR` },
+      lines, `report the conflict to the controller; on its word merge ${ctx.base} into the pin (never rebase), resolve, push, run /ship, then ${Q} ready --pr ${o.pr} --screen|--no-screen`, 'none');
+  }
   if (!loadReady(env.repo.keys, o.pr, head)) rebuildFromNotes(ctx.repoRoot, env.remote);
   const prior = loadReady(env.repo.keys, o.pr, head);
   const slug = repoSlug(ctx);

lib/ship-queue-train.ts

@@ -18,6 +18,7 @@ import { randomBytes } from 'node:crypto';
 import { DISPATCH_INPUTS, awaitNonceRun, dispatchWorkflow, listDispatchRuns, makeNonce, runForNonce, type WorkflowDispatch } from '../scripts/eval-bg';
 import { changelogDamage, mergeChangelogs, parseChangelog, type ChangelogSection } from './changelog-check';
 import type { ResultStatus } from './headless-artifacts';
+import { baseOverlapLine } from './pr-wait';
 import { versionAtRef, versionFiles } from './restamp';
 import type { LoadedPolicy, ShipPolicy, TrainPolicy } from './ship-policy';
 import { loadBasePolicy, type QueueContext } from './ship-queue';
@@ -194,6 +195,39 @@ export function mergeStep(repoRoot: string, pred: string, head: string, policy:
   return 'error' in out ? { conflicts: [`(${out.error})`] } : { tree: out.tree };
 }
 
+export interface BaseOverlap { touched: string[]; conflicts: string[] | null; unknown: boolean; line: string | null }
+
+/**
+ * What gstack-pr-wait's base-overlap event tells a lane, read from git at
+ * READY and at train build (handoff GSTA-119): the files the base changed
+ * since this head forked that the head also changes and, when `verdict` is
+ * asked for, whether the train's own merge of the head onto the base is
+ * clean. A base that moved onto nothing the PR touches is one line of fact; a
+ * conflict is what QUEUE_TRAIN_CONFLICT would otherwise discover at the
+ * scratch merge. A git read that fails is `unknown`, said so, never "no
+ * overlap". Train build asks for no verdict: its own chain merges each pin
+ * onto its prefix, and a merge onto the bare base would contradict it for T2+.
+ */
+export function baseOverlap(repoRoot: string, base: string, baseSha: string, head: string, policy: ShipPolicy, verdict = true): BaseOverlap {
+  const fork = git(repoRoot, ['merge-base', baseSha, head]);
+  const none = { touched: [] as string[], conflicts: null, unknown: false };
+  if (fork.ok && fork.out === baseSha) return { ...none, line: null };
+  const theirs = fork.ok && fork.out ? git(repoRoot, ['diff', '--name-only', fork.out, baseSha]) : fork;
+  const ours = fork.ok && fork.out ? git(repoRoot, ['diff', '--name-only', fork.out, head]) : fork;
+  if (!fork.ok || !fork.out || !theirs.ok || !ours.ok) {
+    const why = [fork, theirs, ours].find(r => !r.ok)?.out.split('\n')[0] || 'git read failed';
+    return { ...none, unknown: true, line: `BASE_OVERLAP: could not compare ${base}@${baseSha.slice(0, 12)} with ${head.slice(0, 12)} (${why}); conflicts unknown` };
+  }
+  const changed = new Set(theirs.out.split('\n').filter(Boolean));
+  const touched = ours.out.split('\n').filter(f => f && changed.has(f)).sort();
+  if (!touched.length) return { ...none, line: `BASE: ${base} moved to ${baseSha.slice(0, 7)} since this PR forked; its new commits touch no file this PR changes` };
+  if (!verdict) return { touched, conflicts: null, unknown: false, line: `BASE_OVERLAP: ${baseOverlapLine(base, baseSha, touched)}` };
+  const m = mergeStep(repoRoot, baseSha, head, policy);
+  const conflicts = 'conflicts' in m ? m.conflicts : null;
+  const said = conflicts ? `the train's merge onto ${base} conflicts beyond the CHANGELOG and the stamp files (${conflicts.join(', ')})` : `the train's merge onto ${base} is clean`;
+  return { touched, conflicts, unknown: false, line: `BASE_OVERLAP: ${baseOverlapLine(base, baseSha, touched)}; ${said}` };
+}
+
 const TRAIN_IDENTITY = { GIT_AUTHOR_NAME: 'gstack-ship-queue', GIT_AUTHOR_EMAIL: 'ship-queue@gstack.local', GIT_COMMITTER_NAME: 'gstack-ship-queue',
   GIT_COMMITTER_EMAIL: 'ship-queue@gstack.local', GIT_AUTHOR_DATE: '2000-01-01T00:00:00Z', GIT_COMMITTER_DATE: '2000-01-01T00:00:00Z' };
 
@@ -353,12 +387,15 @@ async function newTrain(env: TrainEnv, o: BuildOptions): Promise<{ t: TrainRecor
     `ORDER: ${order.train.map((s, i) => `${i + 1}:#${s.pr}(${s.screen.choice === 'screen' ? 'screened' : 'unscreened'})`).join(' ')}${explicit ? ' (explicit --order)' : ''}`,
     ...(order.held.length ? [`HELD: ${order.held.map(s => '#' + s.pr).join(',')} (${order.split ? 'unscreened PRs form their own train' : `over the train size ${env.train.max_size}`})`] : []),
   ];
-  const due = fillDue(order.train, env.train, nowIso(), o.urgent || o.repair);
-  if (due && !o.dryRun) return outPending({ code: 'QUEUE_TRAIN_WAITING', message: `${order.train.length} of ${env.train.max_size} PRs; the oldest READY dispatches at ${due}` }, lines, `${Q} train build ${prsArgs(order.train, !!explicit)} (after ${due}, or with more PRs, or --urgent for one)`, 'none');
   const baseSha = git(env.ctx.repoRoot, ['rev-parse', `${env.remote}/${env.ctx.base}`]).out;
   const key = sha256(JSON.stringify({ base: baseSha, slots: order.train.map(s => `${s.pr}@${s.head}`) }));
-  const live = listTrains(env.repo.keys).find(t => t.key === key && !TERMINAL_STATES.includes(t.state) && t.state !== 'red' && t.state !== 'unattributed');
-  if (live) return { t: Object.assign(live, { held: order.held.map(s => s.pr) }), lines };
+  // The scratch merges run before the fill rule (#3485): a QUEUE_TRAIN_WAITING result is also a clean dry-run merge of
+  // every seated pin, and a pin the base moved onto is named (files only; the chain below is the merge verdict) before the
+  // train is refused for it.
+  for (const s of order.train) {
+    const overlap = baseOverlap(env.ctx.repoRoot, env.ctx.base, baseSha, s.head, env.policy.policy, false);
+    if (overlap.touched.length || overlap.unknown) lines.push(`#${s.pr} ${overlap.line}`);
+  }
   const chain = buildChain(env.ctx.repoRoot, baseSha, order.train, env.policy.policy);
   if ('stoppedAt' in chain) {
     const without = order.train.filter(s => s.pr !== chain.stoppedAt);
@@ -366,6 +403,11 @@ async function newTrain(env: TrainEnv, o: BuildOptions): Promise<{ t: TrainRecor
       `QUEUE_TRAIN_CONFLICT: train ${key.slice(0, 12)} stopped at #${chain.stoppedAt} (conflicts: ${chain.conflicts.join(', ')}) after merging ${chain.merged.map(p => '#' + p).join(',') || '(none)'}; nothing dispatched, no branch pushed; fix: ${Q} train build ${prsArgs(without, !!explicit)}`],
     without.length ? `${Q} train build ${prsArgs(without, !!explicit)}` : 'none', 'none');
   }
+  lines.push(`MERGE_CHECK: T1..T${order.train.length} merge clean onto ${env.ctx.base}@${baseSha.slice(0, 12)} (dry run in object storage; nothing pushed)`);
+  const due = fillDue(order.train, env.train, nowIso(), o.urgent || o.repair);
+  if (due && !o.dryRun) return outPending({ code: 'QUEUE_TRAIN_WAITING', message: `${order.train.length} of ${env.train.max_size} PRs; the oldest READY dispatches at ${due}` }, lines, `${Q} train build ${prsArgs(order.train, !!explicit)} (after ${due}, or with more PRs, or --urgent for one)`, 'none');
+  const live = listTrains(env.repo.keys).find(t => t.key === key && !TERMINAL_STATES.includes(t.state) && t.state !== 'red' && t.state !== 'unattributed');
+  if (live) return { t: Object.assign(live, { held: order.held.map(s => s.pr) }), lines };
   const head = chain.prefixes.at(-1)!;
   const nonce = makeNonce('gate', head.commit, baseSha, randomBytes(3).toString('hex'), { purpose: 'train' });
   const now = nowIso();

lib/ship-receipt-prose.ts

@@ -34,7 +34,9 @@ const ESCAPE = /[.*+?^${}()|[\]\\]/g;
  * is not a `pass` claim), compared with the receipt. A leaf key shared by
  * several counters (`pass` under every gate lane, `passed` under gate and ci)
  * names any of them: the claim matches when its number equals one of those
- * counters, and a mismatch names the first of them.
+ * counters, and a mismatch names the first of them. A number that is the value
+ * of a `key=` or `key:` pair is that key's claim alone, never the start of a
+ * `<n> <name>` claim about the next key (#3513).
  */
 export function checkReceiptProse(receipt: Record<string, unknown>, body: string): ProseCheck {
   const prose = body.replace(BLOCK_RE, '');
@@ -43,12 +45,20 @@ export function checkReceiptProse(receipt: Record<string, unknown>, body: string
     const leaf = field.split('.').at(-1)!;
     byLeaf.set(leaf, [...(byLeaf.get(leaf) ?? []), { field, value }]);
   }
+  const pattern = (leaf: string) => leaf.replace(ESCAPE, '\\$&').replaceAll('_', '[ _-]');
+  // The number-then-name branch must not start on a `key=value` or `key: value` pair's value: in `passes=2 blocking_found=1`
+  // and `passes: 2 blocking_found: 1` the 2 is passes' claim, not a "2 blocking_found" one. The offsets of every counter
+  // key's value are collected first, whatever the whitespace around the separator, and a claim starting there is skipped.
+  const values = new Set<number>();
+  for (const leaf of byLeaf.keys()) for (const m of prose.matchAll(new RegExp(`(?<![\\w-])${pattern(leaf)}\\s*[=:]\\s*(?=\\d)`, 'gi'))) values.add(m.index + m[0].length);
   const mismatches: ProseClaim[] = [];
   let checked = 0;
   for (const [leaf, counters] of byLeaf) {
-    const name = leaf.replace(ESCAPE, '\\$&').replaceAll('_', '[ _-]');
-    const re = new RegExp(`(?:(?<![\\w-])${name}\\s*[=:]\\s*(\\d+)\\b|\\b(\\d+)\\s+${name}(?![\\w-]))`, 'gi');
-    for (const m of prose.matchAll(re)) {
+    const name = pattern(leaf);
+    const re = new RegExp(`(?:(?<![\\w-])${name}\\s*[=:]\\s*(\\d+)\\b|(?<![=:\\w.-])(\\d+)\\s+${name}(?![\\w-]))`, 'gi');
+    for (let m = re.exec(prose); m; m = re.exec(prose)) {
+      // A skipped value hands the scan back right after its digits, so the key that follows still makes its own pair claim.
+      if (m[2] !== undefined && values.has(m.index)) { re.lastIndex = m.index + m[2].length; continue; }
       checked++;
       const n = Number(m[1] ?? m[2]);
       if (!counters.some(c => c.value === n)) mismatches.push({ field: counters[0]!.field, prose: n, receipt: counters[0]!.value });

lib/ship-receipt.ts

@@ -11,6 +11,7 @@ import { approvalSummary } from './approval-citations';
 import { renderIssueLines } from './issue-links';
 import { matchesAny } from './ship-policy';
 import { stampFileKind, versionStampOnly } from './pregate/stamp-only';
+import { stampOnlyLine } from './restamp';
 
 export const RECEIPT_FENCE = 'gstack-ship-receipt';
 const KEY_ORDER: Array<keyof ShipReceipt | 'docs'> = [
@@ -165,22 +166,32 @@ export interface CarryPlanInput {
 export interface CarryRecord { from: string; reason: string; product_patch_id: string; changed_paths: string[]; merge_commits: number; range_diff: string; covers: string[]; reruns: string[] }
 export type CarryPlan = { ok: true; carry: CarryRecord } | { ok: false; code: 'RECEIPT_CARRY_PRODUCT_CHANGED' | 'RECEIPT_CARRY_REFUSED'; message: string };
 
-/** The file's text at the reviewed head and at HEAD, with the release version each head carries (from `VERSION`, else `package.json`). */
-export interface CarryChange { before: string | null; after: string | null; oldVersion?: string | null; newVersion?: string | null }
+/**
+ * The file's text at the reviewed head and at HEAD, with the release version each head carries (from `VERSION`, else
+ * `package.json`) and the base policy's `stamp_patterns` templates.
+ */
+export interface CarryChange { before: string | null; after: string | null; oldVersion?: string | null; newVersion?: string | null; templates?: readonly string[] }
 
 /**
  * A policy-named stamp path's change is a stamp when the new text reads as the
- * old text once every new version in it is the old one: the rule the tree receipt
- * applies to the same `stamp_paths`, `stamp_files` and `release_outputs` before
- * the train accepts a stamp commit. No version change, or a file missing on
- * either side, is not a stamp.
+ * old text once every new version in it is the old one, or when every changed
+ * line is a stamp-form rewrite to the new version (`stampOnlyLine`: the old
+ * version or a newer tentative `*_KEY_VERSION` the slot stamp moved, #3512): the rule
+ * the tree receipt applies to the same `stamp_paths`, `stamp_files` and
+ * `release_outputs` before the train accepts a stamp commit. No version
+ * change, or a file missing on either side, is not a stamp.
  */
 export function stampOnlyText(change: CarryChange): boolean {
   const { before, after, oldVersion, newVersion } = change;
   if (before === null || after === null || !oldVersion || !newVersion || oldVersion === newVersion) return false;
   // Digit-bounded: replacing 1.2.3 inside 11.2.3 or 1.2.34 would read a different version's stamp as this one's.
   const escaped = newVersion.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
-  return after.replace(new RegExp(`(?<![0-9])${escaped}(?![0-9])`, 'g'), oldVersion) === before;
+  if (after.replace(new RegExp(`(?<![0-9])${escaped}(?![0-9])`, 'g'), oldVersion) === before) return true;
+  const was = before.split('\n');
+  const now = after.split('\n');
+  if (was.length !== now.length) return false;
+  const changed = was.map((line, i) => i).filter(i => was[i] !== now[i]);
+  return changed.length > 0 && changed.every(i => stampOnlyLine(was[i]!, now[i]!, oldVersion, newVersion, change.templates ?? []));
 }
 
 /**

lib/tree-receipt.ts

@@ -15,7 +15,7 @@ import * as fs from 'node:fs';
 import * as path from 'node:path';
 import { normalizeChangelog } from './changelog-check';
 import { compareIdentity, executionIdentity, normalizeLockfile, readBundle, type EvidenceBundle, type ExecutionIdentity } from './evidence-bundle';
-import { parseUnifiedDiff, rewriteStamps, versionAtRef, type FileDiff, type GateAheadRecord } from './restamp';
+import { parseUnifiedDiff, stampOnlyLine, versionAtRef, type FileDiff, type GateAheadRecord } from './restamp';
 import type { ResultCodeName } from './result-codes';
 import { matchesAny, type ShipPolicy } from './ship-policy';
 
@@ -49,15 +49,18 @@ function blob(cwd: string, treeish: string, rel: string): string | null {
   return r.status === 0 ? r.stdout : null;
 }
 
-/** Every removed/added pair differs only by the exact stamp rewrites (old -> new). */
+/**
+ * Every removed/added pair differs only by stamp rewrites to the new version:
+ * the old version, or a tentative one (newer than it) the stamp_paths stage
+ * moved (#3512), each in a stamp form (`stampOnlyLine`).
+ */
 function stampOnlyHunks(hunks: FileDiff['hunks'], oldV: string, newV: string, templates: readonly string[]): { why: string; line: number } | null {
   for (const h of hunks) {
     if (h.removed.length !== h.added.length) return { why: `${h.removed.length} line(s) removed, ${h.added.length} added`, line: h.at };
     for (let i = 0; i < h.removed.length; i++) {
       const r = h.removed[i]!;
       const a = h.added[i]!;
-      const rewritten = rewriteStamps(r, oldV, newV, templates);
-      if (rewritten.count === 0 || rewritten.text !== a) return { why: `${JSON.stringify(r.trim().slice(0, 60))} -> ${JSON.stringify(a.trim().slice(0, 60))}`, line: h.at + i };
+      if (!stampOnlyLine(r, a, oldV, newV, templates)) return { why: `${JSON.stringify(r.trim().slice(0, 60))} -> ${JSON.stringify(a.trim().slice(0, 60))}`, line: h.at + i };
     }
   }
   return null;

package.json

@@ -1,6 +1,6 @@
 {
   "name": "gstack",
-  "version": "1.91.277",
+  "version": "1.91.280",
   "description": "Garry's Stack — Claude Code skills + fast headless browser. One repo, one install, entire AI engineering workflow.",
   "license": "MIT",
   "type": "module",

review/sections/review-army.md

@@ -173,20 +173,22 @@ Core findings never earn a specialist confidence boost. Preserve `advisory`,
 
 #### 4. Lead judgment
 
-Reviewers fill the space they are given. Before any merged finding counts, the
-parent reads each one and drops or downgrades it on these grounds, saying so:
-- **Nit gravity.** When a reviewer's whole return is style, naming and preference,
-  the code is probably fine; write that down and count nothing from it as a defect.
+Reviewers fill the space they are given. Before a merged finding counts, the
+parent reads it and drops or downgrades it on these grounds, saying so:
+- **Nit gravity.** A reviewer whose whole return is style, naming and preference
+  found nothing; say so and count no defect.
 - **Hypothetical, not actual.** A defect names the call site or input that reaches
-  it. One that does not is a note until someone traces it.
+  it; one that does not is a note until someone traces it.
 - **Structure for a case that does not exist.** Advice to abstract, generalize or
-  guard a path nothing takes is advisory, never a defect.
+  guard a path nothing takes is advisory.
 - **A different design is not a defect.** "I would have done it differently" with
-  no failing case behind it is dropped.
+  no failing case is dropped.
 - **Missing context.** A finding that assumes a rule the repository documents
   otherwise is dropped with the pointer.
-Agreement between reviewers that did not see each other's output is the strongest
-signal; the `MULTI-SPECIALIST CONFIRMED` boost above records it. A reviewer that
+- **On record.** A class `docs/review-policy.md` rules below the floor is a note,
+  named by its ruling.
+Independent agreement between reviewers is the strongest signal (the
+`MULTI-SPECIALIST CONFIRMED` boost records it). A reviewer that
 returned `NO FINDINGS` was right until a traced defect says otherwise.
 
 #### 5. Apply specialist confidence gates

scripts/eval-readiness.ts

@@ -5,7 +5,8 @@
  * head's paid slices start only after (1) the PR has been quiet for the
  * window (no newer push since this run was created; `vars.EVALS_QUIET_MINUTES`,
  * default 10; the `evals-no-debounce` label sets it to 0) and (2) Free Tests
- * (free-tests.yml) has concluded success on the same head SHA. Selection,
+ * (free-tests.yml) has concluded success on the same head SHA, as GitHub's
+ * pull_request run or as a workflow_dispatch run on the branch. Selection,
  * trials and verdict rules are unchanged: only the start time moves.
  *
  * A newer push cancels the run through the workflow's cancel-in-progress
@@ -40,9 +41,25 @@ export type ReadinessDecision =
   | { kind: 'superseded'; by: string }
   | { kind: 'blocked'; reason: string };
 
-/** The newest Free Tests pull_request run on this exact head decides; a rerun updates that run in place. */
+/**
+ * The Free Tests run events that count as this head's CI, in order of
+ * precedence: GitHub's own pull_request run (it builds the merge ref, the tree
+ * a merge would test), then a workflow_dispatch run on the branch (the
+ * dispatch route in docs/ship-policy.md for a head whose only conflict with
+ * the base is the CHANGELOG, which GitHub builds no merge ref for and so never
+ * runs pull_request CI on; #3508). A dispatched run builds the head alone, so
+ * it decides only for a head with no pull_request run: it never overrides a red
+ * pull_request run (rerun that run), and a lane's extra dispatch never holds a
+ * head its pull_request CI cleared. A push run is a branch build, not this
+ * head's gate.
+ */
+export const FREE_TESTS_EVENTS: readonly string[] = ['pull_request', 'workflow_dispatch'];
+
+/** The newest Free Tests run on this exact head under the first event present decides; a rerun updates that run in place. */
 export function freeTestsState(runs: FreeTestsRun[], headSha: string): FreeTestsState {
-  const newest = runs.filter(run => run.head_sha === headSha && run.event === 'pull_request').sort((a, b) => b.id - a.id)[0];
+  const atHead = runs.filter(run => run.head_sha === headSha);
+  const event = FREE_TESTS_EVENTS.find(e => atHead.some(run => run.event === e));
+  const newest = atHead.filter(run => run.event === event).sort((a, b) => b.id - a.id)[0];
   if (!newest) return { state: 'missing' };
   if (newest.status !== 'completed') return { state: 'running', runId: newest.id };
   if (newest.conclusion === 'success') return { state: 'success', runId: newest.id };
@@ -111,8 +128,9 @@ export function waitForReadiness(o: WaitOptions): { decision: ReadinessDecision;
   if (quietUntil > now()) sleep(quietUntil - now());
   let lastError = '';
   for (;;) {
+    // Every event at the head; freeTestsState keeps the pull_request and workflow_dispatch runs.
     const listed = ghJson<{ workflow_runs?: FreeTestsRun[] }>(gh, ['api', '-X', 'GET', `repos/${o.repo}/actions/workflows/free-tests.yml/runs`,
-      '-f', `head_sha=${o.headSha}`, '-f', 'event=pull_request', '-f', 'per_page=20']);
+      '-f', `head_sha=${o.headSha}`, '-f', 'per_page=20']);
     const pr = ghJson<{ head?: { sha?: string } }>(gh, ['api', `repos/${o.repo}/pulls/${o.pr}`]);
     lastError = !listed.ok ? listed.error : !pr.ok ? pr.error : '';
     if (!lastError && listed.ok && pr.ok) {

scripts/harness-version.json

@@ -17,7 +17,7 @@
     "scripts/lib/published-text.ts": {"blob":"3f2e4d616b5dd0204035710c66b64d49055c5714","since":6},
     "scripts/lib/shard-engine.ts": {"blob":"dc714d7d897aed185c0ec3d2c92ad22d954039e5","since":6,"nonBehavioral":"createShardSandbox drops the Capy launch markers (GSTACK_SESSION_KIND, GSTACK_CAPY, GSTACK_EPHEMERAL, GSTACK_INSTALL_VERDICT) from shard environments; CI never sets them, so no case's input or verdict changes there, and on a Capy machine fixtures now see the session kind they set themselves instead of the installer's"},
     "scripts/test-paid-shards.ts": {"blob":"5439d0a8f6e1569fa7c924af20d67c30c81d5d30","since":6,"nonBehavioral":"credential rejections are classed infra instead of assertion: no case's pass or fail changes; only failed trials whose provider rejected the run's key move class, so no rerun spends on a rejected key"},
-    "scripts/test-pr-profile.ts": {"blob":"a80de69f1923afed67127640d68c99e1a8528c5b","since":6,"nonBehavioral":"classify the new weekly-release.yml workflow and scripts/release-cut.ts as free-lane only PR files: the release cut runs on a cron and manual dispatch after merge and inside no case, so no case's selection on a PR, input or verdict changes"},
+    "scripts/test-pr-profile.ts": {"blob":"188710401a08156f585aebcb699151c60e3d6273","since":6,"nonBehavioral":"#3499: RELEASE_METADATA_FILES exempts NOTICE.md, LICENSE and licenses/** and captain/templates/** joins FREE_ONLY_PR_FILES; selection only, no case runs or is judged differently"},
     "scripts/test-strict-output.ts": {"blob":"01185d20dc9761b04b9b2f751ec98664fe0eaf93","since":6},
     "test/helpers/agent-sdk-runner.ts": {"blob":"210bba92d241a90df578b2c7fc943ac5755d0025","since":6},
     "test/helpers/auq-substance-panel.ts": {"blob":"4881260ef793668e4963a062db6d06a5e1f7a40f","since":6},

scripts/lib/windows-curation.ts

@@ -75,6 +75,7 @@ export const KNOWN_WINDOWS_INCOMPATIBLE: Array<{ file: string; reason: string }>
   { file: 'test/pregate-dart.test.ts', reason: 'its dart, flutter and fvm stand-ins are bash scripts on a colon-joined PATH, which Bun.which and CreateProcess cannot use on Windows, and its containment case needs a POSIX symlink; the Dart lane code it drives is POSIX-run like the other pregate fixtures' },
   { file: 'test/pregate-non-js.test.ts', reason: 'spawns bin/gstack-pregate through test/helpers/pregate-fixture.ts runBin and expects the Linux-host lanes verdict, like test/pregate-docs-lane-review.test.ts and test/pregate-selection.test.ts' },
   { file: 'test/pregate-remote-gbrain.test.ts', reason: 'spawns bin/gstack-pregate through test/helpers/pregate-fixture.ts runBin (shebang via CreateProcess) and runs the gbrain-shaped ci:ubicloud package script through the POSIX lane runner, like test/pregate-remote.test.ts' },
+  { file: 'test/pregate-remote-stamp-only.test.ts', reason: 'expects the Linux-host lanes verdict like test/pregate-remote.test.ts (on Windows the fixture\'s Linux lane is a requires-remote obligation) and skips itself on win32 (describe.skipIf), so the curated lane would run an empty file (#3485, #3486)' },
   // C-pregate: these drive bin/gstack-issue-links and bin/gstack-ship-receipt through
   // test/helpers/pregate-fixture.ts runBin.
   { file: 'test/stamp-templates.test.ts', reason: 'spawns bin/gstack-tree-receipt through test/helpers/restamp-fixture.ts runBin (shebang via CreateProcess)' },

scripts/resolvers/review-army.ts

@@ -218,20 +218,22 @@ Core findings never earn a specialist confidence boost. Preserve \`advisory\`,
 
 #### 4. Lead judgment
 
-Reviewers fill the space they are given. Before any merged finding counts, the
-parent reads each one and drops or downgrades it on these grounds, saying so:
-- **Nit gravity.** When a reviewer's whole return is style, naming and preference,
-  the code is probably fine; write that down and count nothing from it as a defect.
+Reviewers fill the space they are given. Before a merged finding counts, the
+parent reads it and drops or downgrades it on these grounds, saying so:
+- **Nit gravity.** A reviewer whose whole return is style, naming and preference
+  found nothing; say so and count no defect.
 - **Hypothetical, not actual.** A defect names the call site or input that reaches
-  it. One that does not is a note until someone traces it.
+  it; one that does not is a note until someone traces it.
 - **Structure for a case that does not exist.** Advice to abstract, generalize or
-  guard a path nothing takes is advisory, never a defect.
+  guard a path nothing takes is advisory.
 - **A different design is not a defect.** "I would have done it differently" with
-  no failing case behind it is dropped.
+  no failing case is dropped.
 - **Missing context.** A finding that assumes a rule the repository documents
   otherwise is dropped with the pointer.
-Agreement between reviewers that did not see each other's output is the strongest
-signal; the \`MULTI-SPECIALIST CONFIRMED\` boost above records it. A reviewer that
+- **On record.** A class \`docs/review-policy.md\` rules below the floor is a note,
+  named by its ruling.
+Independent agreement between reviewers is the strongest signal (the
+\`MULTI-SPECIALIST CONFIRMED\` boost records it). A reviewer that
 returned \`NO FINDINGS\` was right until a traced defect says otherwise.
 
 #### 5. Apply specialist confidence gates

scripts/test-pr-profile.ts

@@ -235,6 +235,7 @@ export const FREE_ONLY_PR_FILES = [
   'scripts/paid-test-durations.json',
   'scripts/ubicloud/**',
   'capy/skill/**', // The vendored Capy skill (Drive bootstrapper): run by Capy agents, read by no paid case; test/capy-skill-wrapper.test.ts covers it.
+  'captain/templates/**', // The /captain crew templates (CREW.md, merge-ask.md, slot-call.md, …): copied into a Captain's drive by people and Captains, read by no paid case; free test/captain-skill.test.ts covers them (#3499).
   'tsconfig.test.json', // Read only by `tsc -p tsconfig.test.json` (typecheck:test); Bun's runtime reads tsconfig.json.
   '.gitignore', // Changes which untracked files git reports, never tracked content a case reads.
   '.gstack/generated.json', // gstack's own regen registry: gstack-regen/gstack-pregate read it from origin/<base> of the repo they run in; paid cases run in fixture repos.
@@ -285,6 +286,14 @@ export const FREE_ONLY_PR_FILES = [
   '.github/workflows/windows-setup-e2e.yml',
 ] as const;
 
+/**
+ * Release metadata no paid case reads: the third-party notices and licence
+ * texts, and the root LICENSE. A courtesy entry or a licence line never
+ * widens the PR gate (#3499); CHANGELOG.md and VERSION, which the train stamp
+ * rewrites on every slot, are already exempt in knownNonBehaviorFile.
+ */
+export const RELEASE_METADATA_FILES = ['NOTICE.md', 'LICENSE', 'licenses/**'] as const;
+
 /**
  * Inputs every paid case consumes that must always restore the full gate
  * (fail closed, even when a touchfile or a derived reference also names
@@ -317,6 +326,7 @@ export function knownNonBehaviorFile(file: string): boolean {
     // by free doc/generation checks and are not copied into paid fixtures.
     || ['AGENTS.md', 'CLAUDE.md', 'agents-digest/gstack-AGENTS.md'].includes(file)
     || FREE_ONLY_PR_FILES.some(pattern => matchGlob(file, pattern))
+    || RELEASE_METADATA_FILES.some(pattern => matchGlob(file, pattern))
     || (PERIODIC_ONLY_PR_FILES as readonly string[]).includes(file)
     || (file.startsWith('test/fixtures/') && Object.keys(FREE_FIXTURES).some(pattern => matchGlob(file, pattern)))
     // Free test files anywhere (test/, browse/test/, design/test/, make-pdf/test/): no paid case runs or reads a free test.

scripts/train-gate.ts

@@ -26,7 +26,7 @@ import * as fs from 'node:fs';
 import * as path from 'node:path';
 import { normalizeChangelog } from '../lib/changelog-check';
 import { normalizeLockfile } from '../lib/evidence-bundle';
-import { rewriteStamps, versionAtRef } from '../lib/restamp';
+import { stampMentions, rewriteStamps, versionAtRef } from '../lib/restamp';
 import { loadPolicyFromHead, matchesAny, type ShipPolicy } from '../lib/ship-policy';
 import type { BisectPrefixRecord, TrainGateRecord } from '../lib/ship-queue-train-verdict';
 import type { CasePanel } from '../lib/ship-queue-train-record';
@@ -132,20 +132,38 @@ export function fullGateCases(rootDir = ROOT): Map<string, string> {
   return plannedCases(buildRunManifest({ tier: 'gate', profile: 'full', sliceCount: 1, evalsAll: true, env: { EVALS_ALL: '1' }, rootDir }), rootDir);
 }
 
+/**
+ * A release output or stamp path without its stamps: every version mentioned
+ * in a stamp form (the built-in forms and the policy's `stamp_patterns`
+ * templates) collapses to one placeholder. Normalising only the commit's own
+ * version left a tentative `*_KEY_VERSION = '<next minor>'` as written on the
+ * unstamped prefix and normalised once the slot stamp moved it to the released
+ * version, so the two digests differed and train-backstop found no covering
+ * record for main after that train (#3512), and a base-version mention on a
+ * line the base already carried was normalised on the prefix and kept on the
+ * stamped main. Every stamp-form mention collapsing makes the digest blind to
+ * which version a stamp names; the receipt (`stampOnlyLine`) is what judges a
+ * stamp commit's content.
+ */
+export function stampFreeText(text: string, templates: readonly string[]): string {
+  let out = text;
+  for (const v of [...stampMentions(text, templates)].sort((a, b) => b.length - a.length)) out = rewriteStamps(out, v, '\0VERSION', templates).text;
+  return out;
+}
+
 /**
  * A commit's tree modulo stamps: every path with its mode and blob, except the
  * files a restamp rewrites, which contribute their content without the stamp
  * (the version file nothing, package.json without its version, lockfiles
  * without their root version, the CHANGELOG with its headings normalized,
- * release outputs and stamp paths with their own version replaced). Two
- * commits that differ only by a restamp have the same digest.
+ * release outputs and stamp paths with every stamp-form version mention
+ * replaced). Two commits that differ only by a restamp have the same digest.
  */
 export function stampFreeTree(git: Git, commit: string, repoRoot: string, policy: ShipPolicy): string | { error: string } {
   const listed = git(['ls-tree', '-r', '-z', '--full-tree', commit]);
   if (!listed.ok) return { error: `cannot list ${commit.slice(0, 12)} (${listed.out.split('\n')[0]})` };
   const at = versionAtRef(repoRoot, commit);
   const versionRel = 'rel' in at ? at.rel : 'VERSION';
-  const version = 'version' in at ? at.version : null;
   const lines: string[] = [];
   for (const item of listed.out.split('\0').filter(Boolean)) {
     const [meta, rel] = [item.slice(0, item.indexOf('\t')), item.slice(item.indexOf('\t') + 1)];
@@ -161,7 +179,7 @@ export function stampFreeTree(git: Git, commit: string, repoRoot: string, policy
     if (stamp === 'manifest') { try { const parsed = JSON.parse(text); delete parsed.version; text = JSON.stringify(parsed); } catch { /* not JSON: kept verbatim */ } }
     else if (stamp === 'lockfile') text = normalizeLockfile(base, text);
     else if (stamp === 'changelog') text = normalizeChangelog(text);
-    else if (version) text = rewriteStamps(text, version, '\0VERSION', policy.stamp_patterns).text;
+    else text = stampFreeText(text, policy.stamp_patterns);
     lines.push(`${rel}\0${mode}\0${createHash('sha256').update(text).digest('hex')}`);
   }
   return createHash('sha256').update(lines.sort().join('\n')).digest('hex');

scripts/ubicloud/ubi-runner.sh

@@ -590,35 +590,69 @@ cmd_down() {
 }
 
 # pack SRC: write a gzipped checkout tarball to stdout (tracked + untracked
-# files that are not ignored, plus .git; a non-git directory is taken whole).
-# A linked worktree's .git is a one-line pointer into the main repository,
-# useless on the VM, so it is replaced by a self-contained clone of the
-# checked-out branch carrying the worktree's own index.
+# files that are not ignored, plus a .git snapshot; a non-git directory is
+# taken whole). The live .git is never tarred: another process running git in
+# the checkout (a review subagent's diff, a trial merge's worktree add) writes
+# index.lock, packed-refs and objects while tar reads, and GNU tar exits 1
+# on a file that changed under it (GSTA-121's PREGATE_REMOTE_INFRA). Instead
+# the refs are fetched into a fresh repository through git's own transport,
+# which reads a consistent view, the remote.* config is replayed so
+# origin/<base> still resolves, and the index is copied (git writes it
+# atomically through a rename). A linked worktree's .git is a one-line
+# pointer into the main repository, useless on the VM; the same snapshot
+# replaces it, on the worktree's own branch and index. Not carried: submodule
+# repositories (.git/modules) and a partial clone's promisor (the fetch pulls
+# every missing blob first); gstack has neither.
 cmd_pack() {
-  local src tmp="" gitdir branch
+  local src tmp gitdir branch head entry
   src=$(cd "${1:-.}" && pwd)
   if git -C "$src" rev-parse --is-inside-work-tree >/dev/null 2>&1; then
-    if [ -f "$src/.git" ]; then
-      branch=$(git -C "$src" rev-parse --abbrev-ref HEAD)
-      [ "$branch" != HEAD ] || die "pack: $src is a linked worktree on a detached HEAD; check out a branch so the VM gets a self-contained .git"
-      tmp=$(mktemp -d)
-      gitdir=$(cd "$src" && cd "$(git rev-parse --git-common-dir)" && pwd)
-      git clone -q --bare --no-local --single-branch --branch "$branch" "$gitdir" "$tmp/.git"
-      git -C "$tmp/.git" config core.bare false
-      # The worktree's index rides along so the VM's status matches, except a
-      # split index, whose shared half lives only in the original git dir.
-      [ "$(git -C "$src" config --bool core.splitIndex 2>/dev/null)" = true ] \
-        || cp "$(cd "$src" && cd "$(dirname "$(git rev-parse --git-path index)")" && pwd)/index" "$tmp/.git/index"
+    head=$(git -C "$src" rev-parse --verify -q HEAD) || die "pack: $src has no commits; the VM needs a HEAD to fingerprint"
+    branch=$(git -C "$src" symbolic-ref -q --short HEAD || true)
+    tmp=$(mktemp -d)
+    # The path is baked in: a set -e failure unwinds the function before the trap runs, and a local $tmp is unset there.
+    trap "rm -rf $(printf %q "$tmp")" EXIT
+    # No template hooks from a global init.templateDir, and no auto-gc under the tar.
+    git -c gc.auto=0 init -q --bare --template= "$tmp/.git"
+    git -C "$tmp/.git" -c gc.auto=0 fetch -q --no-tags --update-shallow "$src" '+refs/*:refs/*' \
+      || die "pack: could not snapshot the refs of $src"
+    if [ -z "$branch" ]; then
+      branch="detached-${head:0:12}"
+      git -C "$tmp/.git" fetch -q --no-tags --update-shallow "$src" "+HEAD:refs/heads/$branch"
+    fi
+    git -C "$tmp/.git" symbolic-ref HEAD "refs/heads/$branch"
+    git -C "$tmp/.git" config core.bare false
+    gitdir=$(cd "$src" && cd "$(git rev-parse --git-common-dir)" && pwd)
+    while IFS= read -r -d '' entry; do
+      # A key with no value (`[remote "x"] mirror`) is boolean true.
+      case $entry in *$'\n'*) ;; *) entry="$entry"$'\n'true ;; esac
+      git -C "$tmp/.git" config --add "${entry%%$'\n'*}" "${entry#*$'\n'}"
+    done < <(git -C "$gitdir" config --local --null --get-regexp '^(remote|branch)\.' 2>/dev/null || true)
+    # The checkout's index rides along so the VM's status matches, except a
+    # split index, whose shared half lives only in the original git dir. The
+    # index's blobs that HEAD's tree lacks (staged, intent-to-add, the stages
+    # of an unmerged entry) are behind no ref, so they travel as a pack of
+    # their own; without them the VM's write-tree (gstack-wtree) cannot read a
+    # staged file.
+    if [ "$(git -C "$src" config --bool core.splitIndex 2>/dev/null)" != true ]; then
+      cp "$(cd "$src" && cd "$(dirname "$(git rev-parse --git-path index)")" && pwd)/index" "$tmp/.git/index"
+      # Read the copy, not the live index, so the pack matches it; a gitlink's
+      # sha (mode 160000) is a submodule commit this object store lacks.
+      GIT_INDEX_FILE="$tmp/.git/index" git -C "$src" ls-files -s | awk '$1 != "160000" {print $2}' | sort -u \
+        | comm -23 - <(git -C "$src" ls-tree -r HEAD | awk '{print $3}' | sort -u) \
+        | git -C "$src" pack-objects -q --stdout \
+        | git -C "$tmp/.git" unpack-objects -q \
+        || die "pack: could not snapshot the staged blobs of $src"
     fi
     (
       cd "$src"
-      { git ls-files -z -co --exclude-standard; [ -n "$tmp" ] || printf '.git\0'; } \
+      git ls-files -z -co --exclude-standard \
         | while IFS= read -r -d '' f; do
             if [ -e "$f" ] || [ -L "$f" ]; then printf '%s\0' "$f"; fi
           done \
-        | if [ -n "$tmp" ]; then tar --null -T - -c -C "$tmp" .git -zf -; else tar --null -T - -czf -; fi
+        | tar --null -T - -c -C "$tmp" .git -zf -
     )
-    [ -z "$tmp" ] || rm -rf "$tmp"
+    rm -rf "$tmp"; trap - EXIT
   else
     tar -C "$src" -czf - .
   fi

ship/sections/capy-pr.md

@@ -1,6 +1,6 @@
 <!-- AUTO-GENERATED from capy-pr.md.tmpl — do not edit directly -->
 <!-- Regenerate: bun run gen:skill-docs -->
-## Step 19-C: Capy PR body (What changed, Why, Approvals)
+## Step 19-C: Capy PR body (What changed, Why, Blast radius, Approvals)
 
 Read only when Step 19 printed `CAPY_PR: on`. Under Capy a person decides the merge from the PR and
 the READY report, through a Captain, so the body shows what changed in pictures, why, and who

ship/sections/capy-pr.md.tmpl

@@ -1,4 +1,4 @@
-## Step 19-C: Capy PR body (What changed, Why, Approvals)
+## Step 19-C: Capy PR body (What changed, Why, Blast radius, Approvals)
 
 Read only when Step 19 printed `CAPY_PR: on`. Under Capy a person decides the merge from the PR and
 the READY report, through a Captain, so the body shows what changed in pictures, why, and who

ship/sections/review-army.md

@@ -456,20 +456,22 @@ Core findings never earn a specialist confidence boost. Preserve `advisory`,
 
 #### 4. Lead judgment
 
-Reviewers fill the space they are given. Before any merged finding counts, the
-parent reads each one and drops or downgrades it on these grounds, saying so:
-- **Nit gravity.** When a reviewer's whole return is style, naming and preference,
-  the code is probably fine; write that down and count nothing from it as a defect.
+Reviewers fill the space they are given. Before a merged finding counts, the
+parent reads it and drops or downgrades it on these grounds, saying so:
+- **Nit gravity.** A reviewer whose whole return is style, naming and preference
+  found nothing; say so and count no defect.
 - **Hypothetical, not actual.** A defect names the call site or input that reaches
-  it. One that does not is a note until someone traces it.
+  it; one that does not is a note until someone traces it.
 - **Structure for a case that does not exist.** Advice to abstract, generalize or
-  guard a path nothing takes is advisory, never a defect.
+  guard a path nothing takes is advisory.
 - **A different design is not a defect.** "I would have done it differently" with
-  no failing case behind it is dropped.
+  no failing case is dropped.
 - **Missing context.** A finding that assumes a rule the repository documents
   otherwise is dropped with the pointer.
-Agreement between reviewers that did not see each other's output is the strongest
-signal; the `MULTI-SPECIALIST CONFIRMED` boost above records it. A reviewer that
+- **On record.** A class `docs/review-policy.md` rules below the floor is a note,
+  named by its ruling.
+Independent agreement between reviewers is the strongest signal (the
+`MULTI-SPECIALIST CONFIRMED` boost records it). A reviewer that
 returned `NO FINDINGS` was right until a traced defect says otherwise.
 
 #### 5. Apply specialist confidence gates

test/approval-record.test.ts

@@ -76,6 +76,18 @@ describe('gstack-approvals add and render', () => {
     expect(fs.existsSync(path.join(state, 'projects'))).toBe(false);
   });
 
+  test('a GitHub Actions run id after a run word is not a phone number, and the record keeps it (GSTA-121)', () => {
+    const cwd = repo(); const state = tmp('appr-state-');
+    const words = 'Rule-19 grant: re-READY with --no-screen citing screen 40477638211 GREEN at 5619199800a2; gate=40477638210 passed';
+    const r = bin(['add', '--ask', 'Third rule 19 grant on the GREEN screen', '--words', words, '--by', 'GStack Minder', '--scope', 'one merge'], { cwd, env: { GSTACK_STATE_ROOT: state } });
+    expect(r.status, r.err).toBe(0);
+    const file = /^APPROVAL_RECORDED: a-[0-9a-f]{12} (.*\.jsonl)$/m.exec(r.out)![1]!;
+    expect(fs.readFileSync(file, 'utf8')).toContain('screen 40477638211 GREEN');
+    const control = bin(['add', '--ask', 'ok', '--words', 'call me on 40477638211 or +14155550123', '--by', 'g', '--scope', 's'], { cwd, env: { GSTACK_STATE_ROOT: state } });
+    expect(control.status).toBe(3);
+    expect(control.err).toContain('pii.phone.e164');
+  });
+
   test('missing fields, bad outcomes and impossible dates are usage errors', () => {
     const cwd = repo(); const state = tmp('appr-state-');
     expect(bin(['add', '--ask', 'x'], { cwd, env: { GSTACK_STATE_ROOT: state } }).err).toContain('APPROVAL_RECORD_INVALID');

test/eval-readiness.test.ts

@@ -53,6 +53,25 @@ describe('readiness decision', () => {
     expect(freeTestsState([free(3, 'completed', 'cancelled')], HEAD)).toEqual({ state: 'failed', runId: 3, conclusion: 'cancelled' });
   });
 
+  // #3508: a head whose only conflict with main is the CHANGELOG gets no pull_request CI (GitHub has no merge ref to build);
+  // the dispatch route runs Free Tests through workflow_dispatch on the branch, and readiness must read that run.
+  test('a dispatched-only head: the workflow_dispatch Free Tests run at the exact head decides, green or red, newest first; a pull_request run outranks it', () => {
+    const dispatched = (id: number, status: string, conclusion: string | null, over: Partial<FreeTestsRun> = {}) => free(id, status, conclusion, { event: 'workflow_dispatch', ...over });
+    expect(freeTestsState([dispatched(40448356650, 'completed', 'success')], HEAD)).toEqual({ state: 'success', runId: 40448356650 });
+    expect(freeTestsState([dispatched(5, 'completed', 'failure')], HEAD)).toEqual({ state: 'failed', runId: 5, conclusion: 'failure' });
+    expect(freeTestsState([dispatched(5, 'in_progress', null)], HEAD)).toEqual({ state: 'running', runId: 5 });
+    // A dispatched run on another head, or a push run on this one, is not this head's gate.
+    expect(freeTestsState([dispatched(5, 'completed', 'success', { head_sha: NEWER })], HEAD)).toEqual({ state: 'missing' });
+    expect(freeTestsState([free(5, 'completed', 'success', { event: 'push' })], HEAD)).toEqual({ state: 'missing' });
+    // A pull_request run at the head decides when one exists, whatever its age: it built the merge ref, a dispatch built the
+    // head alone, so a newer green dispatch never overrides a red pull_request run (rerun that run instead) and a red or
+    // running extra dispatch never holds a head its pull_request CI cleared (pre-landing review, pass 1).
+    expect(freeTestsState([free(1, 'completed', 'failure'), dispatched(2, 'completed', 'success')], HEAD)).toEqual({ state: 'failed', runId: 1, conclusion: 'failure' });
+    expect(freeTestsState([dispatched(1, 'completed', 'success'), free(2, 'completed', 'failure')], HEAD)).toEqual({ state: 'failed', runId: 2, conclusion: 'failure' });
+    expect(freeTestsState([free(1, 'completed', 'success'), dispatched(2, 'in_progress', null)], HEAD)).toEqual({ state: 'success', runId: 1 });
+    expect(freeTestsState([free(1, 'completed', 'success'), dispatched(2, 'completed', 'failure')], HEAD)).toEqual({ state: 'success', runId: 1 });
+  });
+
   test('ready needs both a green Free Tests on the head and an elapsed window; a moved head is superseded', () => {
     const green = { state: 'success', runId: 7 } as const;
     expect(readinessDecision({ headSha: HEAD, prHead: HEAD, free: green, quietLeftMs: 0 }).kind).toBe('ready');
@@ -80,6 +99,20 @@ describe('waitForReadiness', () => {
     expect(lines.join('\n')).toContain('Free Tests run 5 passed');
   });
 
+  test('a dispatched-only head starts the gate: the runs listing is not narrowed to pull_request (#3508)', () => {
+    const { gh, calls } = fakeGh([{ free: [free(40448356650, 'completed', 'success', { event: 'workflow_dispatch' })] }]);
+    const c = clock(CREATED + 3 * 3600_000);
+    const argsSeen: string[][] = [];
+    const spy: GhRunner = (args, timeout) => { argsSeen.push(args); return gh(args, timeout); };
+    const { decision, lines } = waitForReadiness({ ...base, gh: spy, now: c.now, sleep: c.sleep });
+    expect(decision.kind).toBe('ready');
+    expect(lines.join('\n')).toContain('Free Tests run 40448356650 passed');
+    const listing = argsSeen.find(args => args.some(a => a.endsWith('free-tests.yml/runs')))!;
+    expect(listing).toContain(`head_sha=${HEAD}`);
+    expect(listing.some(a => a.startsWith('event='))).toBe(false);
+    expect(calls.filter(call => call.endsWith('free-tests.yml/runs'))).toHaveLength(1);
+  });
+
   test('a re-run keeps the original creation time, so an elapsed window is not waited again', () => {
     const { gh } = fakeGh([{ free: [free(5, 'completed', 'success')] }]);
     const c = clock(CREATED + 3 * 3600_000);

test/exercised-map.test.ts

@@ -161,7 +161,7 @@ describe('the five recorded variant-c misses on this checkout', () => {
 
 describe('shadow output never changes the real selection', () => {
   test('mode, cases, judges and reasons equal the selection without a shadow', () => {
-    for (const files of [['bin/gstack-config'], ['design/src/evolve.ts'], ['review/SKILL.md.tmpl'], ['lib/fs-utils.ts', 'docs/x.md'], ['ETHOS.md']]) {
+    for (const files of [['bin/gstack-config'], ['design/src/evolve.ts'], ['review/SKILL.md.tmpl'], ['lib/fs-utils.ts', 'docs/x.md'], ['SECURITY.md']]) {
       const withShadow = selectPrProfile({ selectedE2E: [], selectedJudges: [], changedFiles: files });
       const without = selectPrProfile({ selectedE2E: [], selectedJudges: [], changedFiles: files, exercised: null });
       expect(withShadow.shadow, files.join()).toBeDefined();

test/gstack-restamp-queue.test.ts

@@ -469,7 +469,7 @@ describe('gate-ahead (--after), predecessor movement and queue staleness', () =>
     const j = JSON.parse(r.out);
     expect(j).toMatchObject({ status: 'refused', code: 'QUEUE_STALE', expected: { base: assumed.slice(0, 12) } });
     expect(j.actual.base).toBe(repo.git('rev-parse', 'origin/main'));
-    expect(j.recompute).toMatch(/^gstack-restamp --next --base main --expect-base [0-9a-f]{12}$/);
+    expect(j.recompute).toMatch(/^gstack-restamp --next --base main --expect-base [0-9a-f]{12} --pr 2$/);
     expect(repo.git('status', '--porcelain')).toBe('');
     const text = restamp(repo, ['--next', '--pr', '2', '--expect-base', assumed.slice(0, 12)]);
     expect(text.err).toMatch(/base moved.*\(QUEUE_STALE\)\nrecompute: gstack-restamp --next/);
@@ -495,9 +495,38 @@ describe('gate-ahead (--after), predecessor movement and queue staleness', () =>
     const stale = restamp(repo, ['--next', '--pr', '2', '--expect-order', '1,2,3', '--json'], ghShim(repo, { '1': merged, '3': merged }));
     expect(stale.status).toBe(EXIT.refused);
     expect(JSON.parse(stale.out)).toMatchObject({ code: 'QUEUE_STALE', actual: { prs: { '1': 'merged', '3': 'merged' } } });
-    expect(JSON.parse(stale.out).recompute).toContain('--expect-order 1,2,3');
+    // The recompute is the corrected order (merged PRs dropped), never the stale list again.
+    expect(JSON.parse(stale.out).recompute).toMatch(/ --pr 2 --expect-order 2$/);
     const unmerged = restamp(repo, ['--next', '--pr', '2', '--expect-order', '1,4,2', '--json'], ghShim(repo, { '1': merged, '4': open }));
     expect(JSON.parse(unmerged.out)).toMatchObject({ code: 'QUEUE_STALE', actual: { prs: { '4': 'open' } } });
+    expect(JSON.parse(unmerged.out).message).toContain('#4 must merge before this PR stamps');
+    expect(JSON.parse(unmerged.out).recompute).toMatch(/ --expect-order 4,2$/);
+    // A predecessor GitHub calls merged that the local base does not carry yet (the slot call came before the lane fetched)
+    // stays ahead in the recompute: dropping it would let --next allocate from the stale base the version it already took.
+    const unfetched = { state: 'MERGED', mergeCommit: { oid: repo.git('rev-parse', 'pr-b') }, headRefOid: 'y' };
+    const stale2 = restamp(repo, ['--next', '--pr', '2', '--expect-order', '1,4,2', '--json'], ghShim(repo, { '1': merged, '4': unfetched }));
+    expect(JSON.parse(stale2.out)).toMatchObject({ code: 'QUEUE_STALE', actual: { prs: { '4': 'merged-not-in-base' } }, fix: 'git fetch origin main' });
+    expect(JSON.parse(stale2.out).message).toContain('#4 merged but the local main does not carry the merge yet: git fetch origin main, then rerun');
+    expect(JSON.parse(stale2.out).recompute).toMatch(/ --expect-order 4,2$/);
+  });
+
+  // The slot-call lesson (GSTA-123, 2026-10-27): a lane given "remaining order 3487" ran --expect-order 3487 without its own
+  // PR, so every entry read as an unmerged predecessor and the refusal repeated the same list as its recompute.
+  test('--expect-order without the lane\'s own PR: the refusal says so and the recompute lists this PR first, then the remaining order', () => {
+    const repo = fixture();
+    makePr(repo, 'pr-a');
+    makePr(repo, 'pr-b');
+    repo.git('checkout', '-q', 'pr-a');
+    const open = { state: 'OPEN', mergeCommit: null, headRefOid: 'x' };
+    const r = restamp(repo, ['--next', '--pr', '1', '--expect-order', '2,3', '--json'], ghShim(repo, { '2': open, '3': open }));
+    expect(r.status).toBe(EXIT.refused);
+    const j = JSON.parse(r.out);
+    expect(j).toMatchObject({ code: 'QUEUE_STALE', actual: { prs: { '2': 'open', '3': 'open' } } });
+    expect(j.message).toContain('--pr 1 is not in --expect-order, so every listed PR read as a predecessor: list this PR first, then the remaining order');
+    expect(j.recompute).toMatch(/^gstack-restamp --next --base main --expect-base [0-9a-f]{12} --pr 1 --expect-order 1,2,3$/);
+    // The corrected command passes against the same gh answers.
+    expect(restamp(repo, ['--next', '--pr', '1', '--expect-order', '1,2,3', '--dry-run'], ghShim(repo, { '2': open, '3': open })).status).toBe(EXIT.ok);
+    expect(repo.git('status', '--porcelain')).toBe('');
   });
 });
 

test/gstack-ship-receipt-docs.test.ts

@@ -201,6 +201,23 @@ describe('carry follow-ups: git floor, named refusal, branch-touched walk, digit
     expect(change('1.2.34\n', '1.2.44\n')).toBe(false);
   });
 
+  // #3512: the carry classifies a stamp path as metadata only while its change is a stamp; a tentative key version the slot
+  // stamp moved (next minor -> released) was product, so `carry` refused RECEIPT_CARRY_PRODUCT_CHANGED on a stamp commit.
+  test('stampOnlyText accepts a tentative key version moved to the new version, and refuses a released one moved by hand or any other edit', () => {
+    const templates = ["_KEY_VERSION = '{version}'"];
+    const change = (before: string, after: string) => stampOnlyText({ before, after, oldVersion: '1.2.3.0', newVersion: '1.2.4.0', templates });
+    const keys = (docs: string, old: string, extra = '') => `export const DOCS_KEY_VERSION = '${docs}';\nexport const OLD_KEY_VERSION = '${old}';\n${extra}`;
+    expect(change(keys('1.3.0.0', '1.2.2.0'), keys('1.2.4.0', '1.2.2.0'))).toBe(true);
+    expect(change(keys('1.2.3.0', '1.2.2.0'), keys('1.2.4.0', '1.2.2.0'))).toBe(true);
+    expect(change(keys('1.3.0.0', '1.2.2.0'), keys('1.2.4.0', '1.2.4.0'))).toBe(false);
+    expect(change(keys('1.3.0.0', '1.2.2.0'), keys('1.2.4.0', '1.2.2.0', 'export const EXTRA = 1;\n'))).toBe(false);
+    // Without the template the constant is not a stamp form, so the move is a product change.
+    expect(stampOnlyText({ before: keys('1.3.0.0', '1.2.2.0'), after: keys('1.2.4.0', '1.2.2.0'), oldVersion: '1.2.3.0', newVersion: '1.2.4.0' })).toBe(false);
+    // A mention of another width is never a stamp candidate (tentativeMentions), so moving it by hand is a product change.
+    expect(change(keys('1.3.0', '1.2.2.0'), keys('1.2.4.0', '1.2.2.0'))).toBe(false);
+    expect(stampOnlyText({ before: 'left-pad since: 1.2.3\n', after: 'left-pad since: 1.2.4.0\n', oldVersion: '1.2.3.0', newVersion: '1.2.4.0' })).toBe(false);
+  });
+
   test('planCarry names at most CARRY_NAMED_PATHS paths', () => {
     const many = Array.from({ length: CARRY_NAMED_PATHS + 3 }, (_, i) => `src/f${i}.ts`);
     const plan = planCarry({ from: 'a'.repeat(40), head: 'b'.repeat(40), changed: many.map(p => ({ path: p, class: 'product' as const })), productPatchId: { from: 'x', head: 'y' }, testPatchId: { from: '', head: '' }, changedProduct: many, mergeCommits: 0, rangeDiff: '' });

test/helpers/touchfiles-data.ts

@@ -19,6 +19,10 @@
 /**
  * E2E test touchfiles — keyed by testName (the string passed to runSkillTest).
  * Each test lists the file patterns that, if changed, require the test to run.
+ *
+ * ETHOS.md is registered under the /office-hours and /plan-ceo-review cases,
+ * whose prompts quote it; the tier>=3 preamble (Search Before Building) only
+ * names it as a read-on-demand pointer that no paid case asserts on (#3499).
  */
 export const E2E_TOUCHFILES: Record<string, string[]> = {
   'ship-skipped-queued-finding': [ 'lib/outside-review-result.ts', 'lib/gate-outcomes.ts', 'bin/gstack-codex-probe','bin/gstack-state-root.sh', 'lib/state-root.ts', 'package.json', 'bun.lock', '.github/docker/Dockerfile.ci',
@@ -220,7 +224,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
 
     "test/fixtures/plan-scope-target-aw.json",
 
-     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-plan-ceo-plan-mode.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-scope-selection.ts',  'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log', 'bin/gstack-question-preference',   'test/helpers/fake-plan-seed.ts', 'test/helpers/plan-seed-submission.ts',  'test/fixtures/plan-seed-cli.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
+     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'ETHOS.md', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-plan-ceo-plan-mode.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-scope-selection.ts',  'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log', 'bin/gstack-question-preference',   'test/helpers/fake-plan-seed.ts', 'test/helpers/plan-seed-submission.ts',  'test/fixtures/plan-seed-cli.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
     
      'test/fixtures/design-scope-announcement-ao.json',
     'test/fixtures/design-scope-declaration-ak.json',
@@ -310,7 +314,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
     
     "test/fixtures/plan-scope-recovery-av.json",
     "test/fixtures/design-scope-checkpoint-at.json", 
-     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'plan-eng-review/**', 'plan-design-review/**', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-plan-mode-no-op.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-scope-selection.ts',  'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log', 'bin/gstack-question-preference',   'test/helpers/fake-plan-seed.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
+     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'ETHOS.md', 'plan-eng-review/**', 'plan-design-review/**', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-plan-mode-no-op.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-scope-selection.ts',  'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log', 'bin/gstack-question-preference',   'test/helpers/fake-plan-seed.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
     
      'test/fixtures/design-scope-announcement-ao.json',
     'test/fixtures/design-scope-declaration-ak.json',
@@ -332,7 +336,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
   'office-hours-auto-mode':       [ 'ETHOS.md','bin/gstack-state-root.sh', 'lib/state-root.ts', 'test/fixtures/auto-decide-recommendation-361c.json',
     
     'test/fixtures/auto-decide-target-361c.json',
-  'test/fixtures/native-auto-decide-ag.json', 'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json', 'bin/gstack-skill-start', 'bin/gstack-skill-end', 'office-hours/**', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-office-hours-auto-mode.test.ts', 'test/helpers/plan-mode-evidence.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt',
+  'test/fixtures/native-auto-decide-ag.json', 'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json', 'bin/gstack-skill-start', 'bin/gstack-skill-end', 'office-hours/**', 'ETHOS.md', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-office-hours-auto-mode.test.ts', 'test/helpers/plan-mode-evidence.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt',
      'test/fixtures/pty-companion-cli.ts', 'lib/fs-atomic.ts',    'test/helpers/plan-skill-questions.ts', 'test/fixtures/eng-auq-validation-error.json', 'test/fixtures/bash-directory-permission.json', 'test/fixtures/design-tasks-bash-permission.json',  'test/fixtures/read-permission.json',  'test/fixtures/ceo-split-e5-numbered-description-491.json',  'test/helpers/plan-skill-question-events.ts',  'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/skill-census.ts',  'test/helpers/auto-decision-state.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/owned-claude-transcript.ts', 'test/helpers/plan-count-artifacts.ts', 'test/helpers/plan-count-file-permission.ts', 'test/helpers/plan-count-fixture.ts', 'test/helpers/plan-count-pending-exit.ts', 'test/helpers/plan-count-pending-question.ts', 'test/helpers/plan-count-transcript.ts', 'test/helpers/plan-floor-review.ts', 'test/helpers/plan-floor-target.ts', 'test/helpers/plan-scope-selection.ts', 'test/helpers/plan-seed-submission.ts', 'test/helpers/pty-screen.ts', 'bin/gstack-context-recovery'],
   'office-hours-phase4-fork':     [ 'ETHOS.md','bin/gstack-state-root.sh', 'lib/state-root.ts', 'bin/gstack-skill-start', 'bin/gstack-skill-end', 'office-hours/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/question-tuning.ts', 'test/helpers/llm-judge.ts', 'test/skill-e2e-office-hours-phase4.test.ts', 'bin/gstack-context-recovery' ],
   'llm-judge-recommendation':     ['codex/**', 'test/helpers/llm-judge.ts', 'test/llm-judge-recommendation.test.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'codex/SKILL.md.tmpl', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts'],
@@ -350,7 +354,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
      'test/fixtures/autoplan-public-narration-ad.json',
     'test/fixtures/auto-decide-completed-mode-f359.json',
     'test/helpers/plan-count-transcript.ts',  'test/fixtures/plan-count-cross-cwd-ancestry-0bcd.json', 
-     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'bin/gstack-session-kind', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble/generate-preamble-bash.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'plan-ceo-review/**', 'bin/gstack-question-preference', 'bin/gstack-config', 'bin/gstack-slug', 'bin/gstack-remote-identity.sh', 'hosts/claude/hooks/question-preference-hook.ts', 'lib/bin-context.ts', 'lib/remote-identity.ts', 'hosts/claude/hooks/spawned-directive.ts', 'lib/is-conductor.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-auto-decide-preserved.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log',   'test/helpers/fake-plan-seed.ts', 'test/helpers/plan-seed-submission.ts',  'test/fixtures/plan-seed-cli.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
+     'test/fixtures/auto-decide-saved-ai.json', 'test/fixtures/auto-decide-retry-ai.json','bin/gstack-skill-start', 'bin/gstack-skill-end', 'bin/gstack-session-kind', 'scripts/resolvers/question-tuning.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble/generate-preamble-bash.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'plan-ceo-review/**', 'ETHOS.md', 'bin/gstack-question-preference', 'bin/gstack-config', 'bin/gstack-slug', 'bin/gstack-remote-identity.sh', 'hosts/claude/hooks/question-preference-hook.ts', 'lib/bin-context.ts', 'lib/remote-identity.ts', 'hosts/claude/hooks/spawned-directive.ts', 'lib/is-conductor.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/skill-e2e-auto-decide-preserved.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/native-auto-decide.ts',  'test/fixtures/auto-decide-current-declaration-6aef.json',  'test/fixtures/auto-decide-explanatory-mode-043a.json', 'test/fixtures/auto-decide-explanatory-mode-749df.json',  'test/fixtures/auto-decide-structured-77.json', 'test/helpers/auto-decision-state.ts',  'test/fixtures/auto-decide-state-cab3.json', 'bin/gstack-question-log',   'test/helpers/fake-plan-seed.ts', 'test/helpers/plan-seed-submission.ts',  'test/fixtures/plan-seed-cli.ts', 'test/fixtures/native-auto-decide-ag.json',  'test/fixtures/eng-seeded-completion-ai.json', 'test/helpers/plan-count-pending-exit.ts',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**',
     
     'test/helpers/plan-count-fixture.ts', 
      'test/fixtures/pty-companion-cli.ts', 'lib/fs-atomic.ts',  'test/fixtures/auto-decide-mode-selector-749df.json', 'lib/redact-engine.ts', 'lib/redact-patterns.ts', 'test/helpers/ceo-finding-fixture.ts',  'test/helpers/owned-claude-transcript.ts', 'test/helpers/plan-skill-questions.ts', 'test/fixtures/eng-auq-validation-error.json', 'test/fixtures/bash-directory-permission.json', 'test/fixtures/design-tasks-bash-permission.json',  'test/fixtures/read-permission.json',  'test/fixtures/ceo-split-e5-numbered-description-491.json',  'test/helpers/plan-skill-question-events.ts',  'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/skill-census.ts',     'scripts/resolvers/tasks-section.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/plan-count-artifacts.ts', 'test/helpers/plan-count-file-permission.ts', 'test/helpers/plan-count-pending-question.ts', 'test/helpers/plan-floor-review.ts', 'test/helpers/plan-floor-target.ts', 'test/helpers/plan-scope-selection.ts', 'hosts/claude/hooks/hook-log.ts', 'bin/gstack-context-recovery'],
@@ -380,7 +384,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
      'test/fixtures/ceo-mode-pending-submit.json',
      "test/fixtures/plan-count-cross-cwd-ancestry-0bcd.json", 
 
-     "test/fixtures/ceo-mode-colon-at.json", 'plan-ceo-review/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/helpers/ceo-mode-option.ts',  'test/fixtures/ceo-expansion-disposition-77.json',   'test/helpers/plan-count-transcript.ts',  'test/fixtures/autoplan-public-narration-ad.json',  'test/helpers/plan-count-fixture.ts',   'test/fixtures/ceo-prerequisite-n-call.json', 'test/fixtures/eng-prerequisite-77.json', 'test/skill-e2e-plan-ceo-mode-routing.test.ts',  'test/fixtures/ceo-hold-posture-l.json',  'test/fixtures/ceo-mode-labels-l.json',  'test/fixtures/ceo-checkbox-l.screen.txt',  'test/fixtures/ceo-mode-prerequisite-o-calls.json', 'test/fixtures/ceo-mode-prerequisite-q-call.json',  'test/fixtures/ceo-preview-u-call.json', 'test/fixtures/ceo-preview-u-screen.txt',  'test/fixtures/design-preview-v-screen.txt',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/fixtures/ceo-mode-preview-aa-screen.txt', 'test/fixtures/ceo-count-mode-preview-aa-screen.txt',  'test/fixtures/ceo-expansion-auq-ac.json', 'test/helpers/plan-count-pending-question.ts',    'test/fixtures/ceo-barless-submit-ac.json',  'test/fixtures/pending-question-completion-ad.json',  'test/fixtures/ceo-mode-posture-ad.json',  'test/fixtures/ceo-mode-full-ad.json',  'test/fixtures/ceo-prerequisite-ad-v2.json',  'test/fixtures/ceo-hold-posture-ag.json',
+     "test/fixtures/ceo-mode-colon-at.json", 'plan-ceo-review/**', 'ETHOS.md', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/helpers/ceo-mode-option.ts',  'test/fixtures/ceo-expansion-disposition-77.json',   'test/helpers/plan-count-transcript.ts',  'test/fixtures/autoplan-public-narration-ad.json',  'test/helpers/plan-count-fixture.ts',   'test/fixtures/ceo-prerequisite-n-call.json', 'test/fixtures/eng-prerequisite-77.json', 'test/skill-e2e-plan-ceo-mode-routing.test.ts',  'test/fixtures/ceo-hold-posture-l.json',  'test/fixtures/ceo-mode-labels-l.json',  'test/fixtures/ceo-checkbox-l.screen.txt',  'test/fixtures/ceo-mode-prerequisite-o-calls.json', 'test/fixtures/ceo-mode-prerequisite-q-call.json',  'test/fixtures/ceo-preview-u-call.json', 'test/fixtures/ceo-preview-u-screen.txt',  'test/fixtures/design-preview-v-screen.txt',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/fixtures/ceo-mode-preview-aa-screen.txt', 'test/fixtures/ceo-count-mode-preview-aa-screen.txt',  'test/fixtures/ceo-expansion-auq-ac.json', 'test/helpers/plan-count-pending-question.ts',    'test/fixtures/ceo-barless-submit-ac.json',  'test/fixtures/pending-question-completion-ad.json',  'test/fixtures/ceo-mode-posture-ad.json',  'test/fixtures/ceo-mode-full-ad.json',  'test/fixtures/ceo-prerequisite-ad-v2.json',  'test/fixtures/ceo-hold-posture-ag.json',
      "test/fixtures/ceo-hold-commitment-ar.json",
      'test/fixtures/pty-companion-cli.ts', 'lib/fs-atomic.ts',  'test/helpers/ceo-finding-fixture.ts',   'test/helpers/plan-skill-questions.ts', 'test/fixtures/eng-auq-validation-error.json', 'test/fixtures/bash-directory-permission.json', 'test/fixtures/design-tasks-bash-permission.json',  'test/fixtures/read-permission.json',  'test/fixtures/ceo-split-e5-numbered-description-491.json',  'test/helpers/plan-skill-question-events.ts',  'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/skill-census.ts',  'test/helpers/owned-claude-transcript.ts',    'scripts/resolvers/tasks-section.ts',
     'test/fixtures/ceo-expansion-pacing-77.json', 'test/helpers/auto-decision-state.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/native-auto-decide.ts', 'test/helpers/plan-count-artifacts.ts', 'test/helpers/plan-count-file-permission.ts', 'test/helpers/plan-count-pending-exit.ts', 'test/helpers/plan-floor-review.ts', 'test/helpers/plan-floor-target.ts', 'test/helpers/plan-scope-selection.ts', 'test/helpers/plan-seed-submission.ts'],
@@ -417,7 +421,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
   'ship-section-loading':        [ 'ship/**', 'scripts/resolvers/sections.ts', 'scripts/gen-skill-docs.ts', 'test/helpers/auq-sdk-capture.ts', 'test/helpers/session-runner.ts',  'test/skill-e2e-ship-section-loading.test.ts',
       'scripts/resolvers/testing.ts', 'test/helpers/agent-sdk-runner.ts', 'test/helpers/auq-native-capture.ts', 'test/helpers/auto-decision-state.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/hermetic-skill-runtime.ts', 'test/helpers/llm-judge.ts', 'test/helpers/native-auto-decide.ts', 'test/helpers/owned-claude-transcript.ts', 'test/helpers/plan-count-artifacts.ts', 'test/helpers/plan-count-file-permission.ts', 'test/helpers/plan-count-fixture.ts', 'test/helpers/plan-count-pending-exit.ts', 'test/helpers/plan-count-pending-question.ts', 'test/helpers/plan-count-transcript.ts', 'test/helpers/plan-floor-review.ts', 'test/helpers/plan-floor-target.ts', 'test/helpers/plan-scope-selection.ts', 'test/helpers/plan-seed-submission.ts', 'test/helpers/plan-skill-question-events.ts', 'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/plan-skill-questions.ts', 'test/helpers/pty-screen.ts', 'test/helpers/pty-trust-dialog.ts', 'test/helpers/skill-census.ts'],
   'plan-ceo-section-loading':    [ 'ETHOS.md',
-    'test/fixtures/ceo-fill-lifetime.json', 'plan-ceo-review/**',  'scripts/resolvers/sections.ts', 'scripts/gen-skill-docs.ts', 'test/helpers/auq-sdk-capture.ts', 'test/helpers/session-runner.ts',  'test/helpers/ceo-section-loading-fixture.ts', 'test/helpers/ceo-stale-fill-decision.ts',  'test/skill-e2e-plan-ceo-review-section-loading.test.ts', 'test/fixtures/ceo-section-loading-l-report.md', 'test/fixtures/ceo-section-loading-q-report.md', 'test/fixtures/ceo-section-r-rejected-report.md', 'test/fixtures/ceo-section-s-trace-report.md', 'test/fixtures/ceo-section-u-report.md', 'test/fixtures/ceo-section-y-report.md', 'test/fixtures/ceo-section-aa-report.md',  'test/fixtures/sdk-stale-table-ad-v3.json',  'test/fixtures/sdk-ordering-ae.json',  'test/fixtures/sdk-columnar-af.json',  'test/fixtures/sdk-order-b-ag.json',  'test/fixtures/sdk-schedule-continuation-ah.json',  'test/fixtures/sdk-original-order-ai.json',  'test/fixtures/sdk-compact-sequence-aj.json',
+    'test/fixtures/ceo-fill-lifetime.json', 'plan-ceo-review/**', 'ETHOS.md',  'scripts/resolvers/sections.ts', 'scripts/gen-skill-docs.ts', 'test/helpers/auq-sdk-capture.ts', 'test/helpers/session-runner.ts',  'test/helpers/ceo-section-loading-fixture.ts', 'test/helpers/ceo-stale-fill-decision.ts',  'test/skill-e2e-plan-ceo-review-section-loading.test.ts', 'test/fixtures/ceo-section-loading-l-report.md', 'test/fixtures/ceo-section-loading-q-report.md', 'test/fixtures/ceo-section-r-rejected-report.md', 'test/fixtures/ceo-section-s-trace-report.md', 'test/fixtures/ceo-section-u-report.md', 'test/fixtures/ceo-section-y-report.md', 'test/fixtures/ceo-section-aa-report.md',  'test/fixtures/sdk-stale-table-ad-v3.json',  'test/fixtures/sdk-ordering-ae.json',  'test/fixtures/sdk-columnar-af.json',  'test/fixtures/sdk-order-b-ag.json',  'test/fixtures/sdk-schedule-continuation-ah.json',  'test/fixtures/sdk-original-order-ai.json',  'test/fixtures/sdk-compact-sequence-aj.json',
      "test/fixtures/sdk-reported-coordination-ar.md",  "test/fixtures/sdk-ordered-schedule-ar.md",
      'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/tasks-section.ts', 'test/helpers/agent-sdk-runner.ts', 'test/helpers/auq-native-capture.ts', 'test/helpers/auto-decision-state.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/hermetic-skill-runtime.ts', 'test/helpers/llm-judge.ts', 'test/helpers/native-auto-decide.ts', 'test/helpers/owned-claude-transcript.ts', 'test/helpers/plan-count-artifacts.ts', 'test/helpers/plan-count-file-permission.ts', 'test/helpers/plan-count-fixture.ts', 'test/helpers/plan-count-pending-exit.ts', 'test/helpers/plan-count-pending-question.ts', 'test/helpers/plan-count-transcript.ts', 'test/helpers/plan-floor-review.ts', 'test/helpers/plan-floor-target.ts', 'test/helpers/plan-scope-selection.ts', 'test/helpers/plan-seed-submission.ts', 'test/helpers/plan-skill-question-events.ts', 'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/plan-skill-questions.ts', 'test/helpers/pty-screen.ts', 'test/helpers/pty-trust-dialog.ts', 'test/helpers/skill-census.ts'],
   // Data-driven behavioral guard for the 'plan'/'prompt' carves: one case per
@@ -477,7 +481,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
     
     'test/fixtures/plan-floor-routing-361c.json',
 
-    'test/fixtures/ceo-report-permission-fb10.json',  'test/fixtures/plan-floor-permission-fb10.json', 'test/helpers/plan-count-file-permission.ts',  'test/fixtures/plan-edit-cropped-permission-1579.json',  'test/fixtures/plan-count-permission-target-ad-v2.json',  'test/fixtures/design-crop-gutter-ap.json',  'test/fixtures/plan-count-crop-ak.json',  'test/fixtures/plan-count-long-edit-0bcd.json',  'test/fixtures/plan-count-cropped-wrap-6714.json', 'bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/fixtures/forcing-finding-seeds.ts', 'test/skill-e2e-plan-ceo-finding-floor.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-floor-target.ts',  'test/helpers/plan-count-fixture.ts',  'test/helpers/plan-count-artifacts.ts', 
+    'test/fixtures/ceo-report-permission-fb10.json',  'test/fixtures/plan-floor-permission-fb10.json', 'test/helpers/plan-count-file-permission.ts',  'test/fixtures/plan-edit-cropped-permission-1579.json',  'test/fixtures/plan-count-permission-target-ad-v2.json',  'test/fixtures/design-crop-gutter-ap.json',  'test/fixtures/plan-count-crop-ak.json',  'test/fixtures/plan-count-long-edit-0bcd.json',  'test/fixtures/plan-count-cropped-wrap-6714.json', 'bin/gstack-skill-start', 'bin/gstack-skill-end', 'plan-ceo-review/**', 'ETHOS.md', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json', 'test/helpers/hermetic-skill-runtime.ts',  'test/helpers/pty-trust-dialog.ts',  'test/fixtures/forcing-finding-seeds.ts', 'test/skill-e2e-plan-ceo-finding-floor.test.ts',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt', 'test/helpers/plan-floor-target.ts',  'test/helpers/plan-count-fixture.ts',  'test/helpers/plan-count-artifacts.ts', 
      
        
      'test/fixtures/pty-companion-cli.ts', 'lib/fs-atomic.ts',    'test/helpers/plan-skill-questions.ts', 'test/fixtures/eng-auq-validation-error.json', 'test/fixtures/bash-directory-permission.json', 'test/fixtures/design-tasks-bash-permission.json',  'test/fixtures/read-permission.json',  'test/fixtures/ceo-split-e5-numbered-description-491.json',  'test/helpers/plan-skill-question-events.ts',  'test/helpers/plan-skill-question-hook-scope.ts', 'test/helpers/skill-census.ts',  'scripts/resolvers/tasks-section.ts', 'test/helpers/auto-decision-state.ts', 'test/helpers/autoplan-artifact-digest.ts', 'test/helpers/autoplan-artifact-permission.ts', 'test/helpers/autoplan-artifact-recorder.ts', 'test/helpers/dx-selected-navigation.ts', 'test/helpers/e2e-gate.ts', 'test/helpers/eng-cache-writer-decision.ts', 'test/helpers/native-auto-decide.ts', 'test/helpers/owned-claude-transcript.ts', 'test/helpers/plan-count-pending-exit.ts', 'test/helpers/plan-count-pending-question.ts', 'test/helpers/plan-count-transcript.ts', 'test/helpers/plan-scope-selection.ts', 'test/helpers/plan-seed-submission.ts', 'bin/gstack-context-recovery'],
@@ -567,7 +571,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
     'test/fixtures/plan-count-cropped-wrap-6714.json',
     
      "test/fixtures/plan-count-cross-cwd-ancestry-0bcd.json", 
-  'test/fixtures/design-crop-gutter-ap.json', 'bin/gstack-config', 'plan-ceo-review/**', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'bin/gstack-question-preference', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**', 'test/helpers/hermetic-skill-runtime.ts',   'test/helpers/pty-trust-dialog.ts',  'test/helpers/plan-count-transcript.ts',  'test/fixtures/autoplan-public-narration-ad.json', 'test/helpers/plan-count-pending-exit.ts',   'test/helpers/plan-count-artifacts.ts',  'test/helpers/eval-store.ts', 'test/helpers/plan-count-fixture.ts',      'test/fixtures/ceo-prerequisite-n-call.json', 'test/fixtures/eng-prerequisite-77.json', 'test/fixtures/forcing-finding-seeds.ts', 'test/skill-e2e-plan-ceo-split-overflow.test.ts',  'test/fixtures/ceo-checkbox-l.screen.txt', 'test/helpers/plan-count-file-permission.ts',  'test/fixtures/plan-edit-cropped-permission-1579.json',  'test/fixtures/plan-count-edit-permission-t.json',  'test/fixtures/plan-count-owned-permission-v.json', 'test/fixtures/ceo-questionless-w-native.json',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt',  'test/fixtures/plan-count-permission-ac.json', 'test/fixtures/plan-count-permission-ad.json', 'test/fixtures/plan-count-permission-ae.json', 'test/fixtures/plan-count-permission-target-ad-v2.json', 'test/fixtures/plan-count-permission-ah.json',
+  'test/fixtures/design-crop-gutter-ap.json', 'bin/gstack-config', 'plan-ceo-review/**', 'ETHOS.md', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'bin/gstack-question-preference', 'test/helpers/claude-pty-runner.ts', 'test/helpers/pty/**',  'test/fixtures/design-ui-boxed-question.json',  'test/helpers/pty-screen.ts',   'test/fixtures/pty-screen/**', 'test/helpers/hermetic-skill-runtime.ts',   'test/helpers/pty-trust-dialog.ts',  'test/helpers/plan-count-transcript.ts',  'test/fixtures/autoplan-public-narration-ad.json', 'test/helpers/plan-count-pending-exit.ts',   'test/helpers/plan-count-artifacts.ts',  'test/helpers/eval-store.ts', 'test/helpers/plan-count-fixture.ts',      'test/fixtures/ceo-prerequisite-n-call.json', 'test/fixtures/eng-prerequisite-77.json', 'test/fixtures/forcing-finding-seeds.ts', 'test/skill-e2e-plan-ceo-split-overflow.test.ts',  'test/fixtures/ceo-checkbox-l.screen.txt', 'test/helpers/plan-count-file-permission.ts',  'test/fixtures/plan-edit-cropped-permission-1579.json',  'test/fixtures/plan-count-edit-permission-t.json',  'test/fixtures/plan-count-owned-permission-v.json', 'test/fixtures/ceo-questionless-w-native.json',  'test/fixtures/eng-d2-truncated-border-0bcd.json',   'test/fixtures/eng-d1-clipped-elision-1579.json', 'test/fixtures/eng-d2-planning-prelude-4d.json', 'test/fixtures/ceo-approach-z-call.json', 'test/fixtures/ceo-approach-z-screen.txt',  'test/fixtures/plan-count-permission-ac.json', 'test/fixtures/plan-count-permission-ad.json', 'test/fixtures/plan-count-permission-ae.json', 'test/fixtures/plan-count-permission-target-ad-v2.json', 'test/fixtures/plan-count-permission-ah.json',
     
     'test/fixtures/plan-count-crop-ak.json',
     
@@ -618,7 +622,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
   'plan-ceo-review-prosons-cadence':  [ 'ETHOS.md', 'test/helpers/prosons-posture.ts', 
     
     "test/fixtures/plan-scope-recovery-av.json",
-    "test/fixtures/design-scope-checkpoint-at.json", 'plan-ceo-review/**', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'model-overlays/opus-4-7.md', 'test/skill-e2e-plan-prosons.test.ts',
+    "test/fixtures/design-scope-checkpoint-at.json", 'plan-ceo-review/**', 'ETHOS.md', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'model-overlays/opus-4-7.md', 'test/skill-e2e-plan-prosons.test.ts',
     
 
      "scripts/resolvers/preamble/generate-preamble-bash.ts", "scripts/resolvers/preamble/generate-completion-status.ts",
@@ -627,7 +631,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
   'plan-review-prosons-format':       [ 'ETHOS.md', 'test/helpers/prosons-posture.ts', 
     
     "test/fixtures/plan-scope-recovery-av.json",
-    "test/fixtures/design-scope-checkpoint-at.json", 'plan-ceo-review/**', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'model-overlays/opus-4-7.md', 'test/skill-e2e-plan-prosons.test.ts',
+    "test/fixtures/design-scope-checkpoint-at.json", 'plan-ceo-review/**', 'ETHOS.md', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/preamble/generate-ask-user-format.ts', 'scripts/resolvers/preamble.ts', 'scripts/resolvers/preamble-hoist.ts', 'model-overlays/opus-4-7.md', 'test/skill-e2e-plan-prosons.test.ts',
     
 
      "scripts/resolvers/preamble/generate-preamble-bash.ts", "scripts/resolvers/preamble/generate-completion-status.ts",
@@ -764,7 +768,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
     
     "test/fixtures/plan-scope-recovery-av.json",
      "test/fixtures/disabled-dated-record-at.json",
-    'plan-eng-review/**', 'plan-ceo-review/**', 'hosts/claude.ts', 'hosts/define-host.ts',
+    'plan-eng-review/**', 'plan-ceo-review/**', 'ETHOS.md', 'hosts/claude.ts', 'hosts/define-host.ts',
     'scripts/gen-skill-docs.ts', 'scripts/resolvers/index.ts', 'scripts/resolvers/sections.ts',
     'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'scripts/resolvers/constants.ts',
     'bin/gstack-config', 'bin/gstack-codex-probe', 'bin/gstack-review-log', 'bin/gstack-review-read', 'lib/review-evidence.ts', 'bin/gstack-slug', 'bin/gstack-remote-identity.sh', 'bin/gstack-wtree', 'bin/gstack-brain-enqueue', 'test/helpers/session-runner.ts',
@@ -906,7 +910,7 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
     'test/fixtures/outside-async-task-m-events.json',
      'test/fixtures/autoplan-amend-input-77.json',
      'test/fixtures/autoplan-phase-handoff-6714.json','scripts/resolvers/learnings.ts', 'scripts/resolvers/preamble/generate-completion-status.ts', 'scripts/resolvers/preamble/generate-preamble-bash.ts',  'test/fixtures/plan-scope-recovery-av.json',   'test/fixtures/design-scope-checkpoint-at.json',  'scripts/resolvers/composition.ts',   'autoplan/**', 'lib/claude-journal-records.ts', 'lib/claude-owned-journal.ts', 'codex/**', 'bin/gstack-codex-probe', 'scripts/resolvers/review-dashboard.ts', 'scripts/resolvers/plan-gates.ts', 'scripts/resolvers/spec-review.ts', 'scripts/resolvers/outside-voice-steps.ts', 'scripts/resolvers/review-scope.ts', 'scripts/resolvers/outside-voice.ts', 'scripts/resolvers/design.ts', 'test/skill-e2e-autoplan-dual-voice.test.ts', 'bin/gstack-autoplan-snapshot.ts',    'test/fixtures/autoplan/t-ceo-omitted-obligations.json', 'test/fixtures/autoplan/u-ceo-original-loss.json', 'test/fixtures/autoplan/v-ceo-dangling-references.json',
-    'scripts/resolvers/design-doc-discovery.ts', 'bin/gstack-design-doc-find', 'bin/gstack-ceo-plan-find', 'plan-ceo-review/**', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/testing.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts',  'scripts/resolvers/tasks-section.ts', 'test/helpers/autoplan-phase-observer.ts', 'scripts/resolve-codex-generation-model.ts', 'scripts/resolvers/constants.ts'],
+    'scripts/resolvers/design-doc-discovery.ts', 'bin/gstack-design-doc-find', 'bin/gstack-ceo-plan-find', 'plan-ceo-review/**', 'ETHOS.md', 'plan-eng-review/**', 'plan-design-review/**', 'plan-devex-review/**', 'scripts/resolvers/testing.ts', 'scripts/resolvers/preamble/generate-ask-user-format.ts',  'scripts/resolvers/tasks-section.ts', 'test/helpers/autoplan-phase-observer.ts', 'scripts/resolve-codex-generation-model.ts', 'scripts/resolvers/constants.ts'],
 
   // Multi-provider benchmark adapters — live API smoke against real claude/codex/gemini CLIs
   'benchmark-providers-live': ['bin/gstack-model-benchmark', 'test/helpers/providers/**', 'test/helpers/benchmark-runner.ts', 'test/helpers/pricing.ts', 'test/skill-e2e-benchmark-providers.test.ts'],

test/paid-free-boundary.test.ts

@@ -106,6 +106,25 @@ describe('paid/free dependency boundary', () => {
     } finally { fs.rmSync(root, { recursive: true, force: true }); }
   });
 
+  test('release metadata and crew templates never widen the PR gate; ETHOS.md selects only the prompts that quote it (#3499)', () => {
+    // Release metadata no paid case reads keeps the PR profile, alone and beside a mapped prompt (#3499).
+    for (const file of ['NOTICE.md', 'LICENSE', 'licenses/MIT-pstack.txt', 'captain/templates/merge-ask.md']) {
+      expect(fs.existsSync(path.join(ROOT, file)), file).toBe(true);
+      const result = computePaidCaseSelection({ profile: 'pr', env: {}, changedFiles: [file, 'CHANGELOG.md'] });
+      expect(result.coverage?.mode, file).toBe('pr');
+      expect(result.coverage?.unknownFiles, file).toEqual([]);
+      expect(result.selection, file).toEqual({ e2e: [], judges: [] });
+    }
+    // ETHOS.md is read by the /office-hours and /plan-ceo-review prompts, so it selects their cases and judge, not the whole gate.
+    const ethos = computePaidCaseSelection({ profile: 'pr', env: {}, changedFiles: ['ETHOS.md', 'NOTICE.md'] });
+    expect(ethos.coverage?.mode).toBe('pr');
+    expect(ethos.coverage?.unknownFiles).toEqual([]);
+    expect(ethos.selection.e2e).toContain('plan-ceo-review-plan-mode');
+    expect(ethos.selection.e2e).toContain('office-hours-auto-mode');
+    expect(ethos.selection.judges).toEqual(['plan-ceo-review/SKILL.md modes']);
+    expect(ethos.selection.e2e!.length).toBeLessThan(10);
+  });
+
   test('only audited free-only dependencies avoid unknown-dependency fallback', () => {
     for (const file of [...FREE_ONLY_PR_FILES, 'scripts\\test-free-shards.ts']) {
       const result = computePaidCaseSelection({ profile: 'pr', env: {}, changedFiles: [file] });
@@ -114,7 +133,7 @@ describe('paid/free dependency boundary', () => {
       expect(result.selection).toEqual({ e2e: [], judges: [] });
     }
     // Tracked files outside every mapping and outside the derivable directories restore the full gate.
-    for (const file of ['LICENSE', 'SECURITY.md', 'conductor.json', 'NOTICE.md']) {
+    for (const file of ['SECURITY.md', 'conductor.json']) {
       const result = computePaidCaseSelection({ profile: 'pr', env: {}, changedFiles: [FREE_ONLY_PR_FILES[0], file] });
       expect(result.coverage?.mode, file).toBe('full-fallback');
       expect(result.coverage?.unknownFiles).toContain(file);

test/pr-wait.test.ts

@@ -12,7 +12,7 @@ import { tmpdir } from 'node:os';
 import { join } from 'node:path';
 import { spawn, spawnSync } from 'node:child_process';
 import type { GhResult } from '../lib/ci-gate';
-import { DEFAULT_ADVISORY_BOTS, DEFAULT_GATING_BOTS, PR_WAIT_EXIT, Watcher, botLists, parsePrTarget, parsePrWaitArgs, prWaitMain, stateDir, type PrWaitEnv, type PrWaitOptions } from '../lib/pr-wait';
+import { DEFAULT_ADVISORY_BOTS, DEFAULT_GATING_BOTS, DEFAULT_TIMEOUT_MINUTES, PR_WAIT_EXIT, Watcher, botLists, parsePrTarget, parsePrWaitArgs, prWaitMain, stateDir, type PrWaitEnv, type PrWaitOptions } from '../lib/pr-wait';
 import { GhError, runGh } from '../lib/pr-wait-github';
 import { RESULT_CODES } from '../lib/result-codes';
 
@@ -360,7 +360,9 @@ describe('arguments and the gh layer', () => {
     expect(parsePrTarget('abc')).toMatchObject({ error: "'abc' is not a PR number or PR URL" });
     expect(parsePrTarget('0')).toMatchObject({ error: "'0' is not a valid PR number" });
     expect(parsePrTarget('42', 'nope')).toMatchObject({ error: "'nope' is not owner/name" });
-    expect(parsePrWaitArgs(['42', '--until', 'green', '--interval', '2'])).toEqual({ target: '42', until: 'green', interval: 5, timeout: 60 });
+    expect(parsePrWaitArgs(['42', '--until', 'green', '--interval', '2'])).toEqual({ target: '42', until: 'green', interval: 5, timeout: 60, excludes: [] });
+    expect(parsePrWaitArgs(['42', '--exclude', 'quality-gate, canary', '--exclude', 'canary'])).toMatchObject({ excludes: ['quality-gate', 'canary'] });
+    expect(parsePrWaitArgs(['42', '--exclude', ' , '])).toMatchObject({ error: expect.stringContaining('--exclude needs') });
     expect(parsePrWaitArgs(['42', '--timeout', '0'])).toMatchObject({ timeout: 0 });
     expect(parsePrWaitArgs(['42', '--timeout', 'soon'])).toMatchObject({ error: expect.stringContaining('--timeout needs') });
     expect(parsePrWaitArgs(['--until', 'soon'])).toMatchObject({ error: expect.stringContaining('--until needs') });
@@ -421,6 +423,44 @@ describe('pass-1 review fixes: gates agree, bounded waits, bot identity, local f
     expect(lines()).toEqual(['#42 check failed: quality-gate https://ci.example/quality-gate', expect.stringContaining('(PR_WAIT_BLOCKED;')]);
   });
 
+  // #3495 follow-up (ruled by GStack Minder, 2026-10-26): gstack-ci-gate --exclude has a watcher equivalent, so a non-required
+  // check stuck in pending (a canary, a deploy preview) no longer holds --until green for the whole timeout.
+  test('--exclude leaves a named non-required check out of the verdict: pending does not hold green, failed does not block; a required name is refused and gates', () => {
+    const excluded = (script: Script, ...names: string[]) => {
+      const e = env(fakeGh(script));
+      return new Watcher(e, REPO, 42, 'green', 20, botLists(e.vars), DEFAULT_TIMEOUT_MINUTES, names);
+    };
+    const checks = ok(checksJson(['free-tests', 'pass'], ['windows-free-tests', 'pass'], ['quality-gate', 'pending']));
+    const held = excluded({ rulesets: RULESETS, checks }, 'other-check');
+    expect(pollUntilDone(held, 4)).toBeNull();
+    expect(lines()).toEqual([]);
+    const w = excluded({ rulesets: RULESETS, checks }, 'quality-gate');
+    expect(pollUntilDone(w)).toBe(PR_WAIT_EXIT.met);
+    expect(lines()).toEqual(['#42 green on aaaaaaa: 2 pass; required 2/2 green; no greptile or macroscope yet; 0 unresolved threads; merge state CLEAN; 0 commits behind main']);
+    expect(stderr.join('')).toContain('excluded from the verdict: quality-gate');
+    stdout = []; stderr = [];
+    const red = excluded({ rulesets: RULESETS, checks: ok(checksJson(['free-tests', 'pass'], ['windows-free-tests', 'pass'], ['quality-gate', 'fail'])) }, 'quality-gate');
+    expect(pollUntilDone(red)).toBe(PR_WAIT_EXIT.met);
+    expect(lines()).toEqual([expect.stringMatching(/^#42 green on aaaaaaa: 2 pass;/)]);
+    stdout = []; stderr = [];
+    const refused = excluded({ rulesets: RULESETS, checks: ok(checksJson(['free-tests', 'fail'], ['windows-free-tests', 'pass'])) }, 'free-tests');
+    expect(refused.poll()).toBe(PR_WAIT_EXIT.blocked);
+    expect(lines()[0]).toBe('#42 required check failed: free-tests https://ci.example/free-tests');
+    expect(stderr.join('')).toContain('exclude refused for required check free-tests: a required check gates');
+    stdout = []; stderr = [];
+    // When a required-check source cannot be read the set may be short, so nothing is excluded (gstack-ci-gate refuses its
+    // overrides the same way): the excluded failure gates and the first poll says why.
+    const unknown = excluded({ rulesets: err('gh: HTTP 403: rate limited', 1), checks: ok(checksJson(['free-tests', 'pass'], ['windows-free-tests', 'pass'], ['quality-gate', 'fail'])) }, 'quality-gate');
+    expect(unknown.poll()).toBe(PR_WAIT_EXIT.blocked);
+    expect(lines()[0]).toBe('#42 check failed: quality-gate https://ci.example/quality-gate');
+    expect(stderr.join('')).toContain('exclude refused for quality-gate: the required checks could not be read, so every check gates');
+    stdout = []; stderr = [];
+    // Through main: the flag reaches the watcher.
+    const e = env(fakeGh({ rulesets: RULESETS, checks }));
+    expect(prWaitMain(['42', '--repo', REPO, '--until', 'green', '--exclude', 'quality-gate'], e)).toBe(PR_WAIT_EXIT.met);
+    expect(lines()).toEqual([expect.stringMatching(/^#42 green on aaaaaaa: 2 pass;/)]);
+  });
+
   test('a failed bot check is a failed check, gating or advisory, as gstack-ci-gate counts it; never a completed review', () => {
     const gating = watcher({ pr: prJson({ reviews: [] }), checks: ok(checksJson(['build', 'pass'], ['greptile-review', 'fail'])) });
     expect(pollUntilDone(gating)).toBe(PR_WAIT_EXIT.blocked);

test/pregate-allowlist.test.ts

@@ -84,17 +84,17 @@ describe('allowlist: the secrets check', () => {
     expect(secrets(shape).status).toBe(EXIT.fail);
   });
 
-  // Value: protects=a category entry covers an advisory pii hit (a run id read as a phone number) and the WARN count drops;
+  // Value: protects=a category entry covers an advisory pii hit (a run id after a generic word, which the engine's run-word exemption does not cover, read as a phone number) and the WARN count drops;
   //   fails_when=category matching is id-only; why_new=new; seam=none; no_claim=none
   test('a category entry silences the advisory pii finding for a run id', () => {
     const repo = fixture({ ...BASE_PREGATE, allowlist: [{ finding: 'pii', pattern: '\\d{11}', reason: 'GitHub Actions run ids in reports' }] });
-    commit(repo, { 'docs/runs.md': 'run 40302077688 passed\n' });
+    commit(repo, { 'docs/runs.md': 'job 40302077688 passed\n' });
     expect(secrets(repo).status).toBe(EXIT.ok);
     const r = row(repo);
     expect([r.status, r.code]).toEqual(['pass', undefined]);
     expect(r.detail).toContain('1 allowlisted (pii.phone.e164)');
     const control = fixture(BASE_PREGATE);
-    commit(control, { 'docs/runs.md': 'run 40302077688 passed\n' });
+    commit(control, { 'docs/runs.md': 'job 40302077688 passed\n' });
     secrets(control);
     expect([row(control).status, row(control).code]).toEqual(['warn', 'PREGATE_SECRETS']);
     expect(row(control).detail).toContain('pii.phone.e164');

test/pregate-import-graph-paths.test.ts

@@ -30,6 +30,24 @@ describe('import graph: route-group / dynamic-segment paths and tsconfig path al
     expect(testsFor(g, 'app/[[...slug]]/page.tsx').tests).toEqual(['test/pages.test.ts']);
   });
 
+  test('a relative literal a test reads through new URL or readFileSync maps from the file that names it', () => {
+    const repo = fixture();
+    commit(repo, {
+      'ship/sections/pr-body.md.tmpl': '## Summary\n',
+      'ship/sections/other.md.tmpl': '## Other\n',
+      'docs/guide.md': '# guide\n',
+      'test/publication.test.ts': "import { readFileSync } from 'node:fs';\nreadFileSync(new URL('../ship/sections/pr-body.md.tmpl', import.meta.url), 'utf8');\nreadFileSync('./../docs/guide.md', 'utf8');\n",
+      'test/nested/deep.test.ts': "import { readFileSync } from 'node:fs';\nreadFileSync('../../ship/sections/other.md.tmpl', 'utf8');\n",
+    });
+    const g = buildGraph(repo.work);
+    expect(testsFor(g, 'ship/sections/pr-body.md.tmpl').tests).toEqual(['test/publication.test.ts']);
+    expect(testsFor(g, 'docs/guide.md').tests).toEqual(['test/publication.test.ts']);
+    expect(testsFor(g, 'ship/sections/other.md.tmpl').tests).toEqual(['test/nested/deep.test.ts']);
+    // A relative literal that climbs out of the repository names nothing.
+    commit(repo, { 'test/escape.test.ts': "import { readFileSync } from 'node:fs';\nreadFileSync('../../../ship/sections/pr-body.md.tmpl', 'utf8');\n" });
+    expect(testsFor(buildGraph(repo.work), 'ship/sections/pr-body.md.tmpl').tests).toEqual(['test/publication.test.ts']);
+  });
+
   test('a relative import of a bracketed path already maps (control: the import scanner is not the bug)', () => {
     const repo = fixture();
     commit(repo, {

test/redact-context-exemptions.test.ts

@@ -225,6 +225,14 @@ describe("pii.phone.e164: GitHub Actions run and job ids in their id positions",
       "https://github.com/garrytan/gstack/actions/runs/37235771700/job/111534615007",
       "GITHUB_RUN_ID=37235771700",
       '{"run_id": 37235771700}',
+      // An 11-digit id after a run word, as a READY block, a ship receipt or an approval record names one (GSTA-121's `screen <id> GREEN`).
+      "citing screen 40477638211 GREEN at 5619199800a2",
+      "gate=40477638211",
+      "rerun passed 40477638211",
+      "run #40477638211",
+      "Screens: 40477638211",
+      "dispatched 40477638211 at the pin",
+      "cleared by run 19368712345",
     ],
     [
       "call 37235771700 tomorrow",
@@ -232,6 +240,12 @@ describe("pii.phone.e164: GitHub Actions run and job ids in their id positions",
       "https://example.com/contact/37235771700",
       '{"phone_id": 37235771700}',
       "run 4155550123 laps",
+      // A run word before a 10-digit national number, a bare 11-digit number and a generic word exempt nothing.
+      "gate 4155550123",
+      "the id is 40477638211",
+      "job 40477638211",
+      "workflow 40477638211",
+      "40477638211 failed",
     ],
   );
 });

test/ship-queue-train-ops.test.ts

@@ -157,6 +157,8 @@ describe('#3465: the screen is planned on a scratch merge when GitHub\'s merge r
     expect([dry.status, dry.err]).toEqual([EXIT.ok, '']);
     const main = repo.git('rev-parse', 'origin/main');
     expect(line(dry.out, 'SCREEN_MERGE:')).toMatch(new RegExp(`^SCREEN_MERGE: scratch merge [0-9a-f]{12} of #1@${head.slice(0, 12)} onto main@${main.slice(0, 12)} \\(refs/pull/1/merge [0-9a-f]{12} merges onto [0-9a-f]{12}, not main@${main.slice(0, 12)}\\); would push screen/1-${head.slice(0, 12)} \\(dry run\\)$`));
+    // gstack-pr-wait's base-overlap event, read from git at READY (handoff GSTA-119): main's stamp touched the CHANGELOG this PR also changes.
+    expect(line(dry.out, 'BASE_OVERLAP:')).toBe(`BASE_OVERLAP: main moved to ${main.slice(0, 7)} and touches files in this PR: CHANGELOG.md; the train's merge onto main is clean`);
     expect(repo.git('ls-remote', 'origin', 'refs/heads/screen/*')).toBe('');
     const real = queue(fx, ['ready', '--pr', '1', '--pregate', pregate, '--screen']);
     expect([real.status, real.err]).toEqual([EXIT.ok, '']);
@@ -195,6 +197,15 @@ describe('#3465: the screen is planned on a scratch merge when GitHub\'s merge r
     expect(conflict.err).toMatch(/#1 conflicts with main@[0-9a-f]{12} beyond the CHANGELOG and the stamp files \(pr-1\.txt\); GitHub's merge ref lags too \(origin has no refs\/pull\/1\/merge\).*\(QUEUE_SCREEN_STALE_REF\)/);
     expect(next(conflict.out)).toBe('NEXT: resolve the conflict with main, push, run /ship, then gstack-ship-queue ready --pr 1 --screen');
     expect(dispatches(fx, 'evals.yml').length).toBe(1);
+    const main2 = repo.git('rev-parse', 'origin/main');
+    expect(line(conflict.out, 'BASE_OVERLAP:')).toBe(`BASE_OVERLAP: main moved to ${main2.slice(0, 7)} and touches files in this PR: CHANGELOG.md, pr-1.txt; the train's merge onto main conflicts beyond the CHANGELOG and the stamp files (pr-1.txt)`);
+    // Without a screen there is no scratch merge to refuse at, so READY itself refuses the conflict the train would stop at;
+    // the fix follows the crew template's rule 14 (the base is merged on the controller's word, never by the lane alone).
+    const noScreen = queue(fx, ['ready', '--pr', '1', '--pregate', pregate, '--no-screen']);
+    expect(noScreen.status).toBe(EXIT.refused);
+    expect(noScreen.err).toMatch(new RegExp(`#1 conflicts with main@${main2.slice(0, 12)} beyond the CHANGELOG and the stamp files \\(pr-1\\.txt\\); the train would stop at this PR.*\\(QUEUE_TRAIN_CONFLICT\\)`));
+    expect(next(noScreen.out)).toBe('NEXT: report the conflict to the controller; on its word merge main into the pin (never rebase), resolve, push, run /ship, then gstack-ship-queue ready --pr 1 --screen|--no-screen');
+    expect(fs.existsSync(path.join(fx.state, 'ship-queue', 'ready', repoKey(fx), `1-${head}.json`))).toBe(true);
   }, 120_000);
 });
 

test/ship-queue-train.test.ts

@@ -16,6 +16,7 @@ import { spawnSync } from 'node:child_process';
 import * as fs from 'node:fs';
 import * as path from 'node:path';
 import { EXIT } from '../lib/headless-artifacts';
+import { baseOverlap } from '../lib/ship-queue-train';
 import { TRAIN_HELP } from '../lib/ship-queue-train-cli';
 import { BASE_POLICY, cleanup, runBin } from './helpers/restamp-fixture';
 import { coordinatorClone } from './helpers/ship-queue-fixture';
@@ -352,6 +353,44 @@ describe('train build', () => {
     expect(dispatches(fx)).toEqual([]);
   }, 120_000);
 
+  // Handoff GSTA-119 / #3485: a pooled pin the base moved onto is named with the files, and the conflict is found while the
+  // train is still filling instead of at dispatch time; a base move onto files the PR leaves alone is silent.
+  test('a pin whose files main moved onto gets the base-overlap line; a conflict with main is QUEUE_TRAIN_CONFLICT at slot 1 even while the train fills', () => {
+    const fx = make();
+    const h1 = openPr(fx, '1', { file: 'same.txt', text: 'one\n' });
+    const h2 = openPr(fx, '2');
+    readyReceipt(fx, '1', h1, { at: '2026-10-18T11:00:00.000Z' });
+    readyReceipt(fx, '2', h2, { at: '2026-10-18T11:00:00.000Z' });
+    const { repo } = fx;
+    repo.git('checkout', '-q', '-B', 'mainsim', 'origin/main');
+    fs.writeFileSync(path.join(repo.work, 'same.txt'), 'main\n');
+    fs.writeFileSync(path.join(repo.work, 'unrelated.txt'), 'main\n');
+    repo.git('add', '-A');
+    repo.git('commit', '-qm', 'main moves onto same.txt');
+    repo.git('push', '-q', 'origin', 'HEAD:refs/heads/main');
+    repo.git('checkout', '-q', 'pr-1');
+    const main = repo.git('rev-parse', 'origin/main');
+    const r = queue(fx, ['train', 'build', '--prs', `2@${h2},1@${h1}`]);
+    expect(r.status).toBe(EXIT.fail);
+    // At build the line is the files alone: the chain merges each pin onto its prefix, and that merge is the verdict.
+    expect(line(r.out, '#1 BASE_OVERLAP:')).toBe(`#1 BASE_OVERLAP: main moved to ${main.slice(0, 7)} and touches files in this PR: same.txt`);
+    expect(r.out).not.toContain('#2 BASE');
+    expect(line(r.out, 'QUEUE_TRAIN_CONFLICT:')).toMatch(new RegExp(`stopped at #1 \\(conflicts: same\\.txt\\) after merging (\\(none\\)|#2); nothing dispatched, no branch pushed; fix: gstack-ship-queue train build --prs 2@${h2}$`));
+    expect(r.out).not.toContain('MERGE_CHECK:');
+    expect(remoteRefs(fx)).not.toContain('refs/heads/train/');
+    expect(dispatches(fx)).toEqual([]);
+    const without = queue(fx, ['train', 'build', '--prs', `2@${h2}`]);
+    expect(without.status).toBe(EXIT.refused);
+    expect(without.err).toContain('(QUEUE_TRAIN_WAITING)');
+    expect(line(without.out, 'MERGE_CHECK:')).toBe(`MERGE_CHECK: T1..T1 merge clean onto main@${main.slice(0, 12)} (dry run in object storage; nothing pushed)`);
+    // A git read that fails is never "no overlap" (pre-landing review, pass 1): the comparison is reported as unknown.
+    const unknown = baseOverlap(repo.work, 'main', main, 'f'.repeat(40), TRAIN_POLICY as any);
+    expect(unknown).toMatchObject({ unknown: true, touched: [], conflicts: null });
+    expect(unknown.line).toMatch(new RegExp(`^BASE_OVERLAP: could not compare main@${main.slice(0, 12)} with ffffffffffff \\(.+\\); conflicts unknown$`));
+    // At build the overlap line carries the files alone; the verdict is the chain's.
+    expect(baseOverlap(repo.work, 'main', main, h1, TRAIN_POLICY as any, false).line).toBe(`BASE_OVERLAP: main moved to ${main.slice(0, 7)} and touches files in this PR: same.txt`);
+  }, 120_000);
+
   test("GSTA-67: a PR whose CHANGELOG drops a released heading (#3345's pin 1eb7092) is a CHANGELOG conflict, never merged with the damage", () => {
     const fx = make();
     const h1 = openPr(fx, '1');
@@ -423,6 +462,8 @@ describe('train build', () => {
     expect(filling.status).toBe(EXIT.refused);
     expect(filling.err).toMatch(/1 of 5 PRs; the oldest READY dispatches at 2026-10-18T14:00:00\.000Z.*\(QUEUE_TRAIN_WAITING\)/);
     expect(result(filling.out)).toBe('GSTACK_RESULT: skill=ship-queue status=incomplete run=none');
+    // #3485: the scratch merges run before the fill rule, so a WAITING result also proves the seated pins merge clean.
+    expect(line(filling.out, 'MERGE_CHECK:')).toBe(`MERGE_CHECK: T1..T1 merge clean onto main@${fx.repo.git('rev-parse', 'origin/main').slice(0, 12)} (dry run in object storage; nothing pushed)`);
     fx.write(d => { d.rules['evals.yml'] = GREEN_GATE; });
     const urgent = queue(fx, ['train', 'build', '--prs', `3@${h3}`, '--urgent']);
     expect(urgent.status).toBe(EXIT.ok);

test/ship-receipt-prose.test.ts

@@ -48,6 +48,28 @@ describe('checkReceiptProse', () => {
     expect(bad.mismatches).toEqual([{ field: 'gate.unit.pass', prose: 40, receipt: 41 }]);
     expect(checkReceiptProse(receipt, 'Gate passed: 41. CI passed=41.').mismatches).toEqual([]);
   });
+
+  // #3513: the canonical REVIEW_LOOP line /ship and `gstack-ship-queue ready` print. Before the fix the `<n> <name>` branch
+  // started on passes' value and read `2 blocking_found` as a blocking_found claim of 2 (PROSE: mismatch ... body=2 receipt=1).
+  test('a key=value pair\'s value is that key\'s claim alone: the REVIEW_LOOP line parses as five claims that all match', () => {
+    const loop = { ...receipt, review_loop: { passes: 2, blocking_found: 1, deferred: 0, noted: 4, fixed: 8, scope: 'frozen' } };
+    const line = 'REVIEW_LOOP: passes=2 blocking_found=1 deferred=0 noted=4 fixed=8 scope=frozen';
+    const ok = checkReceiptProse(loop, line);
+    expect(ok).toEqual({ checked: 5, mismatches: [] });
+    expect(proseLine(ok)).toBe('PROSE: 5 receipt counter claims in the body, all match');
+    const one = { ...receipt, review_loop: { passes: 1, blocking_found: 0, deferred: 0, noted: 0, fixed: 0, scope: 'frozen' } };
+    expect(checkReceiptProse(one, 'REVIEW_LOOP: passes=1 blocking_found=0 deferred=0 noted=0 fixed=0 scope=frozen')).toEqual({ checked: 5, mismatches: [] });
+    // A real disagreement on the same line is still caught, and named for the key that carries it.
+    expect(checkReceiptProse(loop, 'REVIEW_LOOP: passes=2 blocking_found=3 deferred=0 noted=4 fixed=8 scope=frozen').mismatches)
+      .toEqual([{ field: 'review_loop.blocking_found', prose: 3, receipt: 1 }]);
+    // Prose claims in the number-then-name spelling keep working when they do not follow a pair.
+    expect(checkReceiptProse(loop, 'The loop ran 2 passes and found 1 blocking_found item.')).toEqual({ checked: 2, mismatches: [] });
+    // The `key: value` spelling with whitespace is the same pair (pre-landing review, pass 1): the value never starts the
+    // next key's claim, while a non-counter word before the number (`Review: 2 passes`) leaves the claim as it was.
+    expect(checkReceiptProse(loop, 'REVIEW_LOOP: passes: 2 blocking_found: 1 deferred: 0 noted: 4 fixed: 8')).toEqual({ checked: 5, mismatches: [] });
+    expect(checkReceiptProse(loop, 'Gate passed: 41 failed: 0')).toEqual({ checked: 2, mismatches: [] });
+    expect(checkReceiptProse(loop, 'Review: 2 passes; passes = 2 fixed = 8.')).toEqual({ checked: 3, mismatches: [] });
+  });
 });
 
 describe('gstack-ship-receipt read --check-prose', () => {

test/train-gate.test.ts

@@ -224,6 +224,40 @@ describe('prefix trees modulo stamps', () => {
     expect(stampFreeTree(git, 'f'.repeat(40), repo.work, policy)).toHaveProperty('error');
   });
 
+  // #3512: a stamp path carrying a tentative `*_KEY_VERSION = '<next minor>'` (gstack's own convention since #3511) digests the
+  // same before and after the slot stamp moves it, so train-backstop finds the record that covers main; a base-version mention
+  // on a line the base already carried digests the same at the prefix (where it equals VERSION) and on the stamped main.
+  test('stampFreeTree collapses every stamp-form mention, not only the commit\'s version: a moved tentative key and a kept released one both digest equal', () => {
+    const git = gitIn(repo.work);
+    const policy = { ...loadPolicyFromHead(repo.work).policy, stamp_paths: ['docs/**/*.md', 'lib/policy.ts'], stamp_patterns: ["_KEY_VERSION = '{version}'"] };
+    const policyFile = (text: string) => { fs.mkdirSync(path.join(repo.work, 'lib'), { recursive: true }); fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), text); };
+    // The unstamped prefix: VERSION 1.2.3.0, a new key written as the tentative 1.3.0.0, a released key at 1.2.3.0 (the base's version).
+    repo.git('checkout', '-q', '-B', 'key-prefix', sha.main!);
+    policyFile("export const DOCS_KEY_VERSION = '1.3.0.0';\nexport const REVIEW_KEY_VERSION = '1.2.3.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\n");
+    repo.git('add', '-A'); repo.git('commit', '-q', '-m', 'prefix with a tentative key');
+    const prefix = repo.git('rev-parse', 'HEAD');
+    // Main after the slot stamp: VERSION 1.2.4.0, the tentative moved to the released version, the released key as written.
+    fs.writeFileSync(path.join(repo.work, 'VERSION'), '1.2.4.0\n');
+    policyFile("export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const REVIEW_KEY_VERSION = '1.2.3.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\n");
+    repo.git('commit', '-qam', 'stamp 1.2.4.0');
+    const stamped = repo.git('rev-parse', 'HEAD');
+    expect(stampFreeTree(git, stamped, repo.work, policy)).toBe(stampFreeTree(git, prefix, repo.work, policy));
+    // A byte beyond a stamp-form mention still changes the digest.
+    policyFile("export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const REVIEW_KEY_VERSION = '1.2.3.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\nexport const EXTRA = 1;\n");
+    repo.git('commit', '-qam', 'a code byte');
+    expect(stampFreeTree(git, repo.git('rev-parse', 'HEAD'), repo.work, policy)).not.toBe(stampFreeTree(git, prefix, repo.work, policy));
+    // The receipt agrees with the digest on the stamp, and refuses a released key moved by hand (older than the gated version;
+    // a mention equal to it is the base's own version, which the stamp_paths stage moves on a branch-added line).
+    const tree = (current: string) => (buildReceipt({ repoRoot: repo.work, gated: prefix, current, policy, policyLabel: 'fixture' }) as { tree: string; residual: string[] }).tree;
+    expect(tree(stamped)).toBe('same-modulo-stamps');
+    repo.git('checkout', '-q', '-B', 'hand-move', prefix);
+    fs.writeFileSync(path.join(repo.work, 'VERSION'), '1.2.4.0\n');
+    policyFile("export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const REVIEW_KEY_VERSION = '1.2.3.0';\nexport const OLD_KEY_VERSION = '1.2.4.0';\n");
+    repo.git('commit', '-qam', 'stamp plus a released key moved by hand');
+    expect(tree(repo.git('rev-parse', 'HEAD'))).toBe('changed');
+    repo.git('checkout', '-q', 'squash');
+  });
+
   test('T1..Tk come from the train head by first parents; a head that is not a train is refused', () => {
     const git = gitIn(repo.work);
     const policy = loadPolicyFromHead(repo.work).policy;

test/tree-receipt-stamp-sensitive.test.ts

@@ -8,7 +8,8 @@ import { afterEach, describe, expect, test } from 'bun:test';
 import * as fs from 'node:fs';
 import * as path from 'node:path';
 import { EXIT } from '../lib/headless-artifacts';
-import { cleanup, ghShim, makeFixtureRepo, makePr, readFile, recordAndBundle, restamp, runBin, type FixtureRepo } from './helpers/restamp-fixture';
+import { stampOnlyLine } from '../lib/restamp';
+import { BASE_POLICY, cleanup, ghShim, makeFixtureRepo, makePr, readFile, recordAndBundle, restamp, runBin, type FixtureRepo } from './helpers/restamp-fixture';
 
 const repos: FixtureRepo[] = [];
 const fixture = (opts?: Parameters<typeof makeFixtureRepo>[0]) => { const r = makeFixtureRepo({ prefix: 'tree-receipt-', ...opts }); repos.push(r); return r; };
@@ -55,6 +56,55 @@ describe('tree receipt: stamps, stamp-sensitive gates and negative fixtures', ()
     expect(r.out).not.toContain('pr-a.txt');
   });
 
+  // #3512: under the next-minor convention a new `*_KEY_VERSION` constant is written as a tentative future version, which the
+  // stamp_paths stage moves to the released one at the slot. The receipt accepted a stamp_paths pair only as base -> new, so the
+  // moved constant classified the file `changed` and the train refused the stamp commit (QUEUE_STAMP_NOT_STAMP_ONLY).
+  test('a tentative key version a restamp moved is a stamp, like the base version; a line that changes more than the mention is not', () => {
+    const repo = fixture({ policy: { ...BASE_POLICY, stamp_paths: ['docs/**/*.md', 'lib/policy.ts'], stamp_patterns: ["_KEY_VERSION = '{version}'"] } });
+    makePr(repo, 'pr-a', { expects: '1.2.4.0' });
+    fs.mkdirSync(path.join(repo.work, 'lib'), { recursive: true });
+    fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), "export const DOCS_KEY_VERSION = '1.3.0.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\n");
+    repo.git('add', '-A'); repo.git('commit', '-q', '-m', 'a new policy key, tentative next minor');
+    const bundle = path.join(repo.root, 'bundle.json');
+    recordAndBundle(repo, repo.work, LANES, bundle);
+    const gatedHead = repo.git('rev-parse', 'HEAD');
+    expect(restamp(repo, ['--version', '1.2.4.0']).status).toBe(EXIT.ok);
+    // The tentative 1.3.0.0 moved; the released 1.2.2.0 stayed.
+    expect(readFile(repo, 'lib/policy.ts')).toBe("export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\n");
+    repo.git('commit', '-qam', 'stamp 1.2.4.0');
+    const r = receipt(repo, ['--gated', gatedHead, '--bundle', bundle]);
+    expect(r.status).toBe(EXIT.ok);
+    expect(r.out).toMatch(/^version:\s+1\.2\.3\.0 -> 1\.2\.4\.0$/m);
+    expect(r.out).toMatch(/^  stamp   lib\/policy\.ts \(stamp_paths\)$/m);
+    expect(r.out).toMatch(/^TREE: same-modulo-stamps$/m);
+    expect(r.out).toMatch(/^gate-reuse: eligible/m);
+    // Negative: the version moves and the key is renamed on the same line; a tentative value moved by hand to the wrong version.
+    fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), "export const DOCS_CAPS_KEY_VERSION = '1.2.4.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\n");
+    repo.git('commit', '-qam', 'rename the key too');
+    const renamed = receipt(repo, ['--gated', gatedHead, '--bundle', bundle]);
+    expect(renamed.status).toBe(EXIT.fail);
+    expect(renamed.out).toMatch(/^  changed lib\/policy\.ts \(stamp_paths: "export const DOCS_KEY_VERSION = '1\.3\.0\.0';" -> "export const DOCS_CAPS_KEY_VERSION = '1\.2\.4\.0';"\)$/m);
+    repo.git('reset', '-q', '--hard', 'HEAD~1');
+    fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), "export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const OLD_KEY_VERSION = '1.2.4.0';\n");
+    repo.git('commit', '-qam', 'move a released key by hand');
+    const wrong = receipt(repo, ['--gated', gatedHead, '--bundle', bundle]);
+    expect(wrong.status).toBe(EXIT.fail);
+    expect(wrong.out).toMatch(/^TREE: changed \(lib\/policy\.ts\)$/m);
+    // A mention of another width (a 3-digit pin in a 4-digit repo) is never a stamp candidate, so moving it by hand to the
+    // new version is a product change even though it reads newer than the gated head (pre-landing review, pass 1).
+    repo.git('reset', '-q', '--hard', 'HEAD~1');
+    fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), "export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\nexport const PIN_KEY_VERSION = '9.9.9';\n");
+    repo.git('commit', '-qam', 'a 3-digit pin');
+    const pinned = repo.git('rev-parse', 'HEAD');
+    fs.writeFileSync(path.join(repo.work, 'lib', 'policy.ts'), "export const DOCS_KEY_VERSION = '1.2.4.0';\nexport const OLD_KEY_VERSION = '1.2.2.0';\nexport const PIN_KEY_VERSION = '1.2.4.0';\n");
+    repo.git('commit', '-qam', 'move the 3-digit pin by hand');
+    expect(stampOnlyLine("export const PIN_KEY_VERSION = '9.9.9';", "export const PIN_KEY_VERSION = '1.2.4.0';", '1.2.3.0', '1.2.4.0', ["_KEY_VERSION = '{version}'"])).toBe(false);
+    expect(stampOnlyLine('left-pad since: 1.2.3', 'left-pad since: 1.2.4.0', '1.2.3.0', '1.2.4.0')).toBe(false);
+    const width = receipt(repo, ['--gated', pinned, '--bundle', bundle]);
+    expect(width.status).toBe(EXIT.fail);
+    expect(width.out).toMatch(/^TREE: changed \(lib\/policy\.ts\)$/m);
+  });
+
   test('negative fixture: a dependency pin equal to the old version is `changed`, not a stamp, even inside stamp_paths', () => {
     const repo = fixture();
     const g = gated(repo);

test/ubicloud-runner.test.ts

@@ -311,6 +311,43 @@ test('pack ships a linked worktree with a self-contained .git on the checked-out
     } finally { rmSync(d, { recursive: true, force: true }); }
   }, 60_000);
 
+  test('pack snapshots a plain checkout without reading its live .git: remotes resolve, a detached HEAD gets a branch, and git running beside it cannot fail the tar', () => {
+    const d = mkdtempSync(join(tmpdir(), 'ubi-pack-live-'));
+    try {
+      const sh = (cmd: string, cwd: string) => { const r = Bun.spawnSync(['bash', '-c', cmd], { cwd, timeout: 60_000, env: { ...process.env, GIT_AUTHOR_NAME: 't', GIT_AUTHOR_EMAIL: 't@t', GIT_COMMITTER_NAME: 't', GIT_COMMITTER_EMAIL: 't@t' } }); expect(r.exitCode, r.stderr.toString()).toBe(0); return r.stdout.toString(); };
+      sh('git init -q --bare origin.git && git clone -q origin.git src 2>/dev/null && cd src && echo a > a && git add a && git commit -qm one && git push -q origin HEAD:main && echo b > b && git add b && git commit -qm two && git push -q origin HEAD:main && git checkout -q --detach && echo c > c && echo s > staged && git add staged && echo i > inta && git add -N inta', d);
+      const src = join(d, 'src');
+      const pack = `bash ${JSON.stringify(join(DIR, 'ubi-runner.sh'))} pack .`;
+      // Twenty seconds of index.lock, worktree and ref churn in the checkout while pack runs three times.
+      const churn = Bun.spawn(['bash', '-c', 'end=$((SECONDS+20)); while [ $SECONDS -lt $end ]; do git update-index -q --refresh; git worktree add -q ../wt HEAD >/dev/null 2>&1; git worktree remove --force ../wt >/dev/null 2>&1; git update-ref refs/heads/churn HEAD; git update-ref -d refs/heads/churn; done'], { cwd: src, stdout: 'ignore', stderr: 'ignore' });
+      try {
+        for (const i of [1, 2, 3]) sh(`${pack} > ../pack-${i}.tgz`, src);
+      } finally { churn.kill(); }
+      sh('mkdir vm && tar -xzf pack-3.tgz -C vm', d);
+      const vm = join(d, 'vm');
+      expect(lstatSync(join(vm, '.git')).isDirectory()).toBe(true);
+      const head = sh('git rev-parse HEAD', src).trim();
+      expect(sh('git rev-parse HEAD && git rev-parse --abbrev-ref HEAD && git rev-parse origin/main && git config --get remote.origin.url && git status --short', vm).split('\n').filter(Boolean))
+        .toEqual([head, `detached-${head.slice(0, 12)}`, head, join(d, 'origin.git'), ' A inta', 'A  staged', '?? c']);
+      // The staged blob and the intent-to-add entry's empty blob are behind no ref; the snapshot carries them, so the VM can read and write-tree them.
+      expect(sh('git fsck --no-progress 2>&1; git diff --cached --name-only', vm).trim()).toBe('staged');
+      const wtree = (cwd: string) => sh(`bash ${JSON.stringify(join(ROOT, 'bin/gstack-wtree'))}`, cwd).trim();
+      expect(wtree(vm)).toBe(wtree(src));
+      // The snapshot carries no hook path or other local config from the source checkout.
+      expect(sh('git config --local --list', vm)).not.toContain('hooksPath');
+      // A failed pack leaves no object copy behind in TMPDIR, and a branch's upstream rides along with its remote.
+      sh('mkdir ../tmpd && git checkout -q -b topic && git branch -q --set-upstream-to=origin/main topic', src);
+      const broken = Bun.spawnSync(['bash', '-c', `rm -f .git/index && TMPDIR=../tmpd ${pack} > ../pack-broken.tgz; echo rc=$?; ls ../tmpd`], { cwd: src, timeout: 60_000 });
+      expect(broken.stderr.toString()).toContain('.git/index');
+      expect(broken.stdout.toString().trim()).toBe('rc=1');
+      // A gitlink in the index (a submodule commit this object store lacks) is not packed, and a valueless
+      // remote key (`[remote "origin"] mirror`) replays as true.
+      sh('git reset -q && git update-index --add --cacheinfo 160000,1111111111111111111111111111111111111111,sub && printf \'[remote "origin"]\\n\\tmirror\\n\' >> .git/config', src);
+      sh('mkdir vm2 && ' + pack.replace('pack .', 'pack src') + ' | tar -xzf - -C vm2', d);
+      expect(sh('git rev-parse --abbrev-ref @{u} && git config --bool --get remote.origin.mirror && git ls-files -s sub | cut -c1-6', join(d, 'vm2')).split('\n').filter(Boolean)).toEqual(['origin/main', 'true', '160000']);
+    } finally { rmSync(d, { recursive: true, force: true }); }
+  }, 90_000);
+
   test('up records the VM in the scratch file before the create request is sent', () => {
     const runner = read('scripts/ubicloud/ubi-runner.sh');
     const create = runner.indexOf('cli "${args[@]}" >/dev/null 2>"$dir/create.err"');