Files
gstack/lib/approval-citations.ts
cadbd0af15 v1.91.116.0 feat: field list 2 wave B: gstack-ship-queue, approval classes, state-machine review, pregate timing check, update-check main moves (#3235)
* fix(capy): the install trailer says behind=<n> and its fix, so a stale install shows in every relayed trailer (field list 2 item 13, #2378)

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

* fix(capy): a lease records its holder's start identity, so a pid reused after a wake is reclaimed; a skipped session-start upgrade ends fail behind, never ok (GBRA-105, GBRA-106)

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

* fix(lease): stale needs proof (unreadable or malformed identities keep the lease), ps identity is timezone-independent, and reclaim never deletes a concurrent renewal

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

* feat(autoplan): enforced round cap and spend budget, deferred close, public export without machine paths, opt-in durable checkpoints (field list 2 W5-W7)

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

* fix(lease): one directory lock for publish, prune and reclaim; reclaim rereads the record judged stale, so a renewal is never deleted or hidden

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

* fix(lease): the directory lock names its holder; broken only when the holder is dead or its pid reused, unlock removes only its own lock, touch exits 3 under a live holder

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

* fix(lease): breakers serialize on .lock.break and reread the lock before removing it; gstack-skill-start reports a lease the held lock kept out

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

* docs(lease): name the double-crash limit of breaking a dead lock

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

* feat(capy): --for review-voice install profile, no-Drive bootstrap command, reviewer prompt trailer (field list 2 W11)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 89bf556c11)

* test: pin review-voice row, helpers row and the no-Drive fallback in existing install tests

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 7f6c70459a)

* test: the DRIVER template names the reviewer trailer relay (pregate lane)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 2545d9a6d2)

* test(context-budget): capy skill eager ceiling for the review-voice subagent paragraph (W11; ratchet protocol, capture value)

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

* feat(eval-plan): prereg diagnostics dry run (W8) and .gstack/eval-models.json lint (W9), WIP

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit b509096218)

* feat(eval-plan): W8/W9 tests, docs/eval-models.md, troubleshooting anchors, /eval-plan steps

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 9335e45de2)

* test(eval-plan): eval-plan/SKILL.md names only defined codes and accepted flags

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit a7a974d268)

* test(review-voice): the launch line points at the verdict's bun-only directory (#3233's launch contract)

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

* test(third-party-actions): the Aside drive-offer detector matches the product name, not the English idiom (paid run 38988927989 tpa-absent-linux false positive)

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

* fix(autoplan): export refuses artifact paths outside the run before writing; a deferred close keeps earlier undecided findings; reopening CEO grants Eng its rerun (review of #3234)

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

* fix(restamp): gate-ahead keeps the lane CHANGELOG entry when both PRs took the same tentative stamp (GBRA-102 diagnosis fixture)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 43c75a6bb6)

* test(tree-receipt): a real gbrain stamp pair is same-modulo-stamps under gbrain committed policy (W1 acceptance)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 6cf53f9fed)

* feat(ship-policy): gate_command, full_e2e and min_gstack_version; unknown keys stay fail-closed; validate prints introducing versions

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 878a7123d7)

* feat(ship-queue): gstack-ship-queue ready, stamp, verify and gate-ahead (WIP, tests next)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 409a5957ad)

* test(ship-queue): stamp journal and resume, verify from a separate clone, gate-ahead attempts, ready

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit ade554087b)

* docs(ship-policy): Run a shared merge queue; new keys and compatibility rule; troubleshooting anchors; /ship pointer

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 7fc73e8892)

* fix(ship): shared-queue pointer rides the show-effective ship: line (ship skeleton has 2 bytes of parity headroom)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 5b33d4a015)

* test(windows): curate the shared-queue tests out of the Windows subset (shebang spawns)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 8deee14728)

* test(fixtures): register the gbrain stamp-pair fixture as free-only

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 563a07aa03)

* test(ship-queue): receipt digest contract names lib/ship-queue.ts for its pregate lane

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 501637c258)

* feat(pregate): timing check flags elapsed-time bounds and sleep-then-assert in added test lines (W4)

lib/pregate/checks/timing.ts prints TIMING_ASSERTION: <file>:<line> <pattern>
and a TIMING_ASSERTIONS summary; gstack-timing-ok: <reason> in any comment
syntax on the line or the line above escapes one. It warns
(PREGATE_TIMING_ASSERTION): measured on gstack (9 of 47 test-touching
commits, 263 hits in the tree) and gbrain (37 of 95 commits; 31 of 38 sites
of the five timing-flake fixes), blocking would stop a fifth to two fifths
of test changes. Adds the /review checklist rule and the /investigate
timing-flake playbook.

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit a3b86b10b2)

* feat(update-check): an equal VERSION with a moved main is an upgrade, via the compare API (W12)

When VERSION matches but ls-remote's main SHA differs from this checkout's
HEAD, bin/gstack-update-check asks the GitHub compare API (status, ahead_by)
through the existing receipted curl and prints
UPGRADE_AVAILABLE <ver> <ver>+<n> <sha>; a diverged or unreadable compare is
CHECK_FAILED (unknown), and behind/identical is up to date. Never fetches.

The cache gains an IDENTITY line (head, compared remote SHA, upstream): an
outcome replays only for the same HEAD and upstream, a same-remote refresh
reuses it without another compare, and a git checkout ignores a legacy
one-line cache (non-git installs keep the one-line format). The snooze is
keyed by the SHA. gstack-upgrade's question, snooze note and What's New step
handle the four-field line; gstack-skill-end reports it.

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 421e66d2a4)

* fix(pregate): treat planning and evidence markdown as an empty lane

A docs-only diff with no .gstack/pregate.json failed lanes as
PREGATE_NO_LANE. Planning and evidence Markdown is now a docs/no-tests
empty lane. Unmapped source files still fail closed.

Co-authored-by: Nick <surlymo@apache.org>
(cherry picked from commit e8793e59e8)

* fix(pregate): keep test-read markdown on its test lane

Unmarked Markdown that no test depends on is a docs/no-tests empty lane.
A file a test reads, including a fixture, still selects that test through
the import graph. Unmapped source and unreferenced test-tree files still
fail closed.

Co-authored-by: Nick <surlymo@apache.org>
(cherry picked from commit 7aabbddf3d)

* test(pregate): reviewer edges for the docs/no-tests empty lane (#3215)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 2fb3fa8613)

* test(pregate): requires-remote on a gbrain-shaped repo (ci:ubicloud, UBI_OWNER) (W10)

A base pregate.json with gbrain's dependencies and remote_gate
"bun run ci:ubicloud" (a package script whose stub needs UBICLOUD_API_KEY,
names VMs from UBI_OWNER and prints PREGATE_REMOTE). Under 8 cores the test
tier is one obligation naming the command; --remote passes the caller's
UBI_OWNER and key through and clears it on the report; without the key the
gate's own error shows and the obligation stays uncleared. #3190's path
needed no change.

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 4d69ff02df)

* test(pregate): the ship receipt counts the timing check (7 checks)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit cb03ef67e7)

* test(context-budget): gstack-upgrade eager ceiling for the W12 equal-version upgrade question (ratchet protocol, capture value)

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

* feat(approval-classes): gstack-approval-classes and ship-receipt citation enforcement (W2, WIP)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 847c7dae3d)

* feat(approval-classes): tests, docs, /review and /ship wiring (W2)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 269bd5b8b5)

* feat(review): state-machines specialist and gstack-state-matrix round-trip check with a captured #6792 regression (W3)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 1a634979c6)

* fix(review-army): keep the --simplification force flag pinned; approval test timeouts

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 789d392bce)

* test(free-fixtures): register the state-matrix-6792 fixture

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 1bacef296d)

* test(state-matrix): cover state-machines NEVER_GATE in gstack-specialist-stats

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
(cherry picked from commit 859063d56f)

* test(fixtures): register the pregate-timing fixtures as free-only (W4)

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

* fix(ship-queue): full_e2e needs bound gate evidence in ready and verify, CI reads every check-run page and commit status, next: quotes its paths (review #3235 P1-1, P1-5, P2-1)

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

* fix(pregate): timing reads a native diff and never passes an unread one

The timing check's git diff honored diff.external, textconv and
diff.noprefix, so a configured driver hid every added line and the check
passed. Run it with --no-ext-diff --no-textconv and explicit a/ b/
prefixes, and report a git diff that fails or does not finish as
incomplete (PREGATE_INCOMPLETE), which a warn downgrade does not soften.

Review of #3235, sibling of P1-4 (lib/pregate/checks/timing.ts:189).

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

* fix(approval-classes): only affirmative APPROVED evidence counts, explicit tokens are authoritative, native diffs fail closed (review #3235 P1-2, P1-3, P1-4)

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

* fix(state-matrix): detect reads the native diff, ignoring diff.external and textconv (review #3235 P1-4)

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

* test(e2e): review-army-state-matrix behavior case on the #6792 fixture (field list 2, W3)

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

* fix(review-army): tighten the State Machines dispatch line so /ship stays under its parity size ratio with main's #3231 review-army text

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

* fix(approval-classes): only top-level approval lines count; quoted, fenced, indented-code and commented text never does (review #3235 r2)

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

* fix(approval-classes): approval evidence is the first non-blank line only, at column 0 (review #3235 r3)

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

* test(autoplan-budget): bind the reservation expectation to lib/autoplan-budget's default estimate (test-of-test ratchet)

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

* fix: keep gstack-upgrade under its size budget, exempt state-matrix's fetch hint, fit the periodic plan's 37 slices

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

* fix(evals): the docs observer accepts the spawned-prefixed gstack-skill-start --help read

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

* fix(windows): curate capy-review-voice out of the Windows subset (POSIX launcher contract)

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

* docs: changelog entry for field list 2 wave A

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

* chore: stamp v1.91.115.0

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

* test: the scope-signal clarity check follows helper-printed signals; the hermetic update-check fixture writes the identity cache

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

* refactor: move the release-path result codes to lib/result-codes-release.ts (module-size ratchet); fit the periodic plan's 38 slices

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

* docs: changelog entry for field list 2 wave B

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

* chore: stamp v1.91.116.0

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

* test: copy lib/result-codes-release.ts wherever a fixture copies lib/result-codes.ts; the scope-signal check requires the helper's computed emission

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

* test: state-matrix stats test resolves its slug under its own state root (home containment)

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

* fix(windows): curate pregate-docs-lane-review out of the Windows subset (runBin shebang, like the other pregate fixtures)

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

* fix(windows): curate pregate-remote-gbrain out of the Windows subset (runBin shebang, like pregate-remote)

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

---------

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Nick <surlymo@apache.org>
2026-10-17 03:32:16 +00:00

229 lines
15 KiB
TypeScript

/**
* approval-citations — the `Approved by <owner> <YYYY-MM-DD> via <ref>`
* lines that let a change in a declared approval class ship (field list 2,
* W2; lib/approval-classes.ts finds the classes). A citation is resolved,
* never trusted as typed:
*
* - a plan sha (7-64 hex) must match the plan of an autoplan run passed with
* `--approval-run <dir>` whose gate is APPROVED (status complete, the p1
* decision answered and not reject/interrogate), with plan.md matching its
* manifest hash;
* - a GitHub comment or review URL must be written by a ship-policy
* maintainer whose login is the cited owner, and the evidence is the body's
* first non-blank line only (`approvalLine`: "Approve" or "Approved" at
* column 0, optionally in `**`, no list marker, quote or indent; nothing
* later in the body counts); a review must also be APPROVED;
* - anything else, or a ref that cannot be read here, is `unverified`.
*
* Resolution alone is not enough: the evidence (the approved plan text or
* the comment's approval line) must bind to the class's current effects, by
* carrying the scope token `<class>@<digest12>` or by naming the class and
* every effect path. A token is authoritative: evidence naming the class with
* another digest approved earlier effects and never falls back to paths. An
* authentic approval of other work, or one given before an effect was added,
* does not bind. An unverified citation counts only as an explicit
* applicability attestation: `for <class>@<digest12>` on the line, matching
* the current scope, so a later effect change invalidates it.
* Caller: bin/gstack-ship-receipt write (APPROVAL_CITATION_MISSING).
*/
import { spawnSync } from 'node:child_process';
import * as fs from 'node:fs';
import * as path from 'node:path';
import { runApprovalScan, type ApprovalHit, type ApprovalScope } from './approval-classes';
import { sha256File } from './headless-artifacts';
import { isUnavailable, loadPolicyFromBase } from './ship-policy';
export interface Citation { raw: string; owner: string; date: string; ref: string; for: Array<{ class: string; digest?: string }> }
export type Resolution =
| { state: 'resolved'; evidence: string; text: string; approvalLine?: true }
| { state: 'rejected'; why: string }
| { state: 'unresolved'; why: string };
export type GhRunner = (args: string[]) => { ok: boolean; out: string; err: string };
export interface CitationContext { runDirs: string[]; maintainers: string[]; gh: GhRunner }
const CLASS_REF = '[a-z][a-z0-9-]*(?:@[0-9a-f]{12,64})?';
const CITATION_RE = new RegExp(`Approved by @?([A-Za-z0-9][\\w.-]*) (\\d{4}-\\d{2}-\\d{2}) via (\\S+)(?: for (${CLASS_REF}(?:,${CLASS_REF})*))?`, 'g');
function validDate(d: string): boolean {
const t = Date.parse(`${d}T00:00:00Z`);
return !Number.isNaN(t) && new Date(t).toISOString().slice(0, 10) === d;
}
/** Every well-formed citation line in `text` (a PR body or --approval values); a bad calendar date is not a citation. */
export function parseCitations(text: string): Citation[] {
const out: Citation[] = [];
for (const m of text.matchAll(CITATION_RE)) {
if (!validDate(m[2]!)) continue;
const ref = m[3]!.replace(/[.,;:)\]>]+$/, '');
const forList = (m[4] ?? '').split(',').filter(Boolean).map(entry => {
const [cls, digest] = entry.split('@');
return digest ? { class: cls!, digest } : { class: cls! };
});
out.push({ raw: m[0], owner: m[1]!.replace(/\.$/, ''), date: m[2]!, ref, for: forList });
}
return out;
}
function readJsonl(file: string): Array<Record<string, any>> {
if (!fs.existsSync(file)) return [];
return fs.readFileSync(file, 'utf8').split('\n').filter(Boolean).flatMap(l => { try { return [JSON.parse(l)]; } catch { return []; } });
}
function planGate(dir: string, run: Record<string, any>): Resolution {
const entry = run.artifacts?.['plan.md'];
const planFile = path.join(dir, entry?.path ?? 'plan.md');
if (!entry || !fs.existsSync(planFile) || sha256File(planFile) !== entry.sha256) return { state: 'rejected', why: `run ${run.run}: plan.md does not match its manifest hash` };
const gate = run.gate ?? {};
const rev = gate.gate_rev ?? run.gate_rev;
const p1 = readJsonl(path.join(dir, 'decisions.jsonl')).find(d => d.label === 'P1' && d.gate_rev === rev);
const approved = run.status === 'complete' && typeof gate.plan === 'string' && !['c', 'e'].includes(gate.plan) && p1 && p1.chosen === gate.plan && p1.status !== 'pending';
if (!approved) return { state: 'rejected', why: `run ${run.run} holds this plan but its gate is not APPROVED (status ${run.status}, p1 ${p1?.chosen ?? 'unanswered'})` };
return { state: 'resolved', evidence: `plan ${run.run} gate_rev=${rev} p1=${gate.plan} answered_by=${gate.answered_by ?? p1.answered_by ?? 'unrecorded'}`, text: fs.readFileSync(planFile, 'utf8') };
}
/** A plan sha: the plan.md (or reviewed plan) hash of an approved autoplan run among `runDirs`. */
export function resolvePlan(ref: string, runDirs: string[]): Resolution {
const want = ref.toLowerCase();
let rejected: Resolution | undefined;
for (const dir of runDirs) {
let run: Record<string, any>;
try { run = JSON.parse(fs.readFileSync(path.join(dir, 'run.json'), 'utf8')); } catch { continue; }
const hashes = [run.artifacts?.['plan.md']?.sha256, run.plan_reality?.plan_sha256].filter((h): h is string => typeof h === 'string');
if (!hashes.some(h => h.toLowerCase().startsWith(want))) continue;
const r = planGate(dir, run);
if (r.state === 'resolved') return r;
rejected = r;
}
return rejected ?? { state: 'unresolved', why: runDirs.length ? `no --approval-run directory holds a plan with sha ${ref}` : `plan sha ${ref}: pass the autoplan run with --approval-run <dir>` };
}
/**
* The approval evidence in a comment or review body: its first non-blank line, and only when that line starts at
* column 0 with "Approve" or "Approved" (optionally in `**`), with no list marker, quote or indent; '' otherwise.
* Nothing later in the body counts, so no quoting or nesting can turn a rejection's example into consent.
*/
export function approvalLine(body: string): string {
const first = body.split(/\r?\n/).find(l => l.trim()) ?? '';
return /^(?:\*\*)?approved?\b/i.test(first) ? first : '';
}
const GH_URL = /^https:\/\/github\.com\/([\w.-]+)\/([\w.-]+)\/(?:pull|issues)\/(\d+)#(?:issuecomment-(\d+)|pullrequestreview-(\d+)|discussion_r(\d+))$/;
/** A comment or review URL: authored by a ship-policy maintainer whose login is the cited owner; a review must be APPROVED; only the body's first-line approval is evidence. */
export function resolveGithub(c: Citation, ctx: CitationContext): Resolution {
const m = GH_URL.exec(c.ref)!;
const [, owner, repo, num, issueComment, review, discussion] = m;
const api = issueComment ? `repos/${owner}/${repo}/issues/comments/${issueComment}`
: review ? `repos/${owner}/${repo}/pulls/${num}/reviews/${review}`
: `repos/${owner}/${repo}/pulls/comments/${discussion}`;
const r = ctx.gh(['api', api]);
if (!r.ok) return { state: 'unresolved', why: `gh api ${api} failed: ${r.err.split('\n')[0] || 'no output'}` };
let body: Record<string, any>;
try { body = JSON.parse(r.out); } catch { return { state: 'unresolved', why: `gh api ${api} returned no JSON` }; }
const login = String(body.user?.login ?? '');
if (!ctx.maintainers.length) return { state: 'unresolved', why: 'the base ship policy lists no maintainers, so the author cannot be classified' };
if (!ctx.maintainers.some(x => x.toLowerCase() === login.toLowerCase())) return { state: 'rejected', why: `${c.ref} is by ${login || 'an unknown author'}, not a ship-policy maintainer` };
if (login.toLowerCase() !== c.owner.toLowerCase()) return { state: 'rejected', why: `the citation names ${c.owner} but ${c.ref} is by ${login}` };
if (review && body.state === 'CHANGES_REQUESTED') return { state: 'rejected', why: `${c.ref} requests changes; it is not an approval` };
if (review && body.state !== 'APPROVED') return { state: 'rejected', why: `${c.ref} is a ${String(body.state ?? 'stateless').toLowerCase()} review; only an APPROVED review counts` };
return { state: 'resolved', evidence: `${review ? 'review' : 'comment'} by ${login}`, text: approvalLine(String(body.body ?? '')), approvalLine: true };
}
export function resolveCitation(c: Citation, ctx: CitationContext): Resolution {
if (/^[0-9a-f]{7,64}$/i.test(c.ref)) return resolvePlan(c.ref, ctx.runDirs);
if (GH_URL.test(c.ref)) return resolveGithub(c, ctx);
return { state: 'unresolved', why: `${c.ref} is neither a plan sha nor a GitHub comment or review URL` };
}
/** The digests `text` names for this class (`<class>@<hex>`), lowercased. */
function namedDigests(text: string, cls: string): string[] {
return [...text.toLowerCase().matchAll(new RegExp(`(?:^|[^a-z0-9-])${cls}@([0-9a-f]{12,64})(?![0-9a-f])`, 'g'))].map(m => m[1]!);
}
/** How the evidence names this exact scope: the token, or (when it names no token for the class) the class plus every effect path; null otherwise. */
export function binding(text: string, scope: ApprovalScope): 'token' | 'paths' | null {
const digests = namedDigests(text, scope.class);
if (digests.length) return digests.some(d => scope.digest.startsWith(d)) ? 'token' : null;
const lower = text.toLowerCase();
const named = new RegExp(`(^|[^a-z0-9-])${scope.class}([^a-z0-9-]|$)`).test(lower);
return named && scope.files.every(f => text.includes(f)) ? 'paths' : null;
}
export interface ApprovalRecord {
class: string; scope: string; effects: string[]; citation: 'verified' | 'unverified';
approved_by: string; date: string; via: string; binding: 'token' | 'paths' | 'attested'; evidence: string;
}
export interface MissingApproval { class: string; token: string; files: string[]; reasons: string[] }
function judge(c: Citation, scope: ApprovalScope, res: Resolution): ApprovalRecord | string {
const base = { class: scope.class, scope: scope.token, effects: scope.hits, approved_by: c.owner, date: c.date, via: c.ref };
if (res.state === 'rejected') return `${c.ref}: ${res.why}`;
if (res.state === 'resolved') {
const how = binding(res.text, scope);
if (how) return { ...base, citation: 'verified', binding: how, evidence: res.evidence };
const stale = namedDigests(res.text, scope.class)[0];
if (stale) return `${c.ref}: ${res.evidence} names ${scope.class}@${stale.slice(0, 12)}, not the current ${scope.token}; a token is authoritative, so approval of earlier effects never falls back to paths`;
const where = res.approvalLine ? ' on its first non-blank line, which must start at column 0 with "Approved" or "Approve"' : '';
return `${c.ref}: ${res.evidence} names neither ${scope.token} nor ${scope.class} with every effect path (${scope.files.join(', ')})${where}; an approval of other work, or one given before these effects, does not cover them`;
}
const attested = c.for.find(f => f.class === scope.class);
if (attested?.digest && scope.digest.startsWith(attested.digest)) return { ...base, citation: 'unverified', binding: 'attested', evidence: `unresolved: ${res.why}` };
if (attested?.digest) return `${c.ref}: attested for ${scope.class}@${attested.digest.slice(0, 12)} but the effects are now ${scope.token} (changed since the attestation)`;
return `${c.ref}: unresolved (${res.why}); an unverified citation must attest "for ${scope.token}"`;
}
/** One record per scope: a verified citation first, else an attested unverified one; otherwise the scope is missing, with every candidate's reason. */
export function evaluateApprovals(scopes: ApprovalScope[], citations: Citation[], ctx: CitationContext): { records: ApprovalRecord[]; missing: MissingApproval[] } {
const cache = new Map<string, Resolution>();
const resolve = (c: Citation) => { const key = `${c.ref}\0${c.owner}`; if (!cache.has(key)) cache.set(key, resolveCitation(c, ctx)); return cache.get(key)!; };
const records: ApprovalRecord[] = [];
const missing: MissingApproval[] = [];
for (const scope of scopes) {
const candidates = citations.filter(c => !c.for.length || c.for.some(f => f.class === scope.class));
const judged = candidates.map(c => judge(c, scope, resolve(c)));
const pick = judged.find((j): j is ApprovalRecord => typeof j !== 'string' && j.citation === 'verified') ?? judged.find((j): j is ApprovalRecord => typeof j !== 'string');
if (pick) records.push(pick);
else missing.push({ class: scope.class, token: scope.token, files: scope.files, reasons: judged.length ? judged as string[] : ['no "Approved by" line names this class or applies to every class'] });
}
return { records, missing };
}
export function defaultGh(args: string[]): { ok: boolean; out: string; err: string } {
const r = spawnSync('gh', args, { encoding: 'utf8', timeout: 60_000, maxBuffer: 16 * 1024 * 1024 });
return { ok: r.status === 0, out: r.stdout ?? '', err: (r.stderr ?? '').trim() };
}
export interface ReceiptApprovalRequest { repoRoot: string; base: string; head: string; citationText: string; runDirs: string[]; gh?: GhRunner }
export type ReceiptApproval =
| { kind: 'skip' }
| { kind: 'ok'; approval: Record<string, unknown>; note?: string }
| { kind: 'invalid'; lines: string[] }
| { kind: 'error'; message: string; code?: 'APPROVAL_SOURCE_UNAVAILABLE' }
| { kind: 'missing'; missing: MissingApproval[] };
const reportOnly = (hits: ApprovalHit[]) => hits.map(h => `${h.class} ${h.file}:${h.line} (${h.text})`);
/** The receipt's approval object, or the refusal: classes from origin/<base>, effects at `head`, citations from the body and --approval. */
export function approvalForReceipt(req: ReceiptApprovalRequest): ReceiptApproval {
const outcome = runApprovalScan({ repoRoot: req.repoRoot, base: req.base, head: req.head });
if ('unavailable' in outcome) return { kind: 'ok', approval: { source: outcome.source, status: 'unavailable', detail: outcome.unavailable, classes: [] }, note: `APPROVAL_CLASSES: unavailable (${outcome.unavailable}); approval classes were not checked` };
if ('errors' in outcome) return { kind: 'invalid', lines: outcome.errors.map(e => `${e.message} at ${e.path || '<root>'} (${outcome.source})`) };
if ('error' in outcome) return { kind: 'error', message: outcome.error, ...(outcome.unreadable ? { code: 'APPROVAL_SOURCE_UNAVAILABLE' as const } : {}) };
const { scan, absent } = outcome;
if (absent && !scan.report_only.length) return { kind: 'skip' };
const policy = loadPolicyFromBase(req.repoRoot, req.base);
const maintainers = isUnavailable(policy) ? [] : policy.policy.maintainers;
const { records, missing } = evaluateApprovals(scan.scopes, parseCitations(req.citationText), { runDirs: req.runDirs, maintainers, gh: req.gh ?? defaultGh });
if (missing.length) return { kind: 'missing', missing };
return { kind: 'ok', approval: { source: scan.source, status: 'checked', classes: records, ...(scan.report_only.length ? { report_only: reportOnly(scan.report_only) } : {}) } };
}
/** The SHIP_RECEIPT summary token: verified, unverified:<classes>, none or unchecked. */
export function approvalSummary(approval: Record<string, any>): string {
if (approval.status === 'unavailable') return 'unchecked';
const classes = (approval.classes ?? []) as ApprovalRecord[];
if (!classes.length) return 'none';
const unverified = classes.filter(c => c.citation === 'unverified').map(c => c.class);
return unverified.length ? `unverified:${unverified.join(',')}` : 'verified';
}