@feigi/fleet-ctl 0.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (245) hide show
  1. package/.claude-plugin/plugin.json +9 -0
  2. package/LICENSE +202 -0
  3. package/NOTICE +18 -0
  4. package/agents/fleet-finisher.agent.md +7 -0
  5. package/agents/fleet-implementer-alt.agent.md +210 -0
  6. package/agents/fleet-implementer.agent.md +210 -0
  7. package/agents/fleet-merge-bot.agent.md +7 -0
  8. package/agents/fleet-review-comments.agent.md +40 -0
  9. package/agents/fleet-review-correctness.agent.md +43 -0
  10. package/agents/fleet-review-runner.agent.md +73 -0
  11. package/agents/fleet-review-silent-failure.agent.md +49 -0
  12. package/agents/fleet-review-simplify.agent.md +46 -0
  13. package/agents/fleet-review-snapshot.agent.md +24 -0
  14. package/agents/fleet-review-tests.agent.md +46 -0
  15. package/agents/fleet-review-types.agent.md +41 -0
  16. package/agents/fleet-review-verifier.agent.md +29 -0
  17. package/commands/review-and-fix.md +98 -0
  18. package/commands/run-merge-bot.md +357 -0
  19. package/commands/run-team-help.md +17 -0
  20. package/package.json +23 -0
  21. package/scripts/agent-test-dir-prose.test.mjs +59 -0
  22. package/scripts/agent-test.test.mjs +1636 -0
  23. package/scripts/ambient-git-vars-mjs-prose.test.mjs +354 -0
  24. package/scripts/ambient-git-vars-prose.test.mjs +192 -0
  25. package/scripts/ancestry-check-position-prose.test.mjs +175 -0
  26. package/scripts/apply-ruleset.test.mjs +420 -0
  27. package/scripts/arg-header-probes-prose.test.mjs +393 -0
  28. package/scripts/arg.mjs +600 -0
  29. package/scripts/arg.test.mjs +1354 -0
  30. package/scripts/board-cli.test.mjs +694 -0
  31. package/scripts/board-identity.test.mjs +429 -0
  32. package/scripts/board-liveness-banner.test.mjs +119 -0
  33. package/scripts/board-prev-shape.test.mjs +229 -0
  34. package/scripts/board-tryparse-null.test.mjs +119 -0
  35. package/scripts/board.html +471 -0
  36. package/scripts/board.mjs +2037 -0
  37. package/scripts/board.test.mjs +3086 -0
  38. package/scripts/brief-after-cut-prose.test.mjs +53 -0
  39. package/scripts/candidates-exit3-prose.test.mjs +302 -0
  40. package/scripts/candidates.mjs +674 -0
  41. package/scripts/candidates.test.mjs +2101 -0
  42. package/scripts/caveat-unification-prose.test.mjs +98 -0
  43. package/scripts/ci-comment-rot-prose.test.mjs +2025 -0
  44. package/scripts/ci-completes-premise-prose.test.mjs +257 -0
  45. package/scripts/ci-state-prose.test.mjs +113 -0
  46. package/scripts/ci-state.mjs +761 -0
  47. package/scripts/ci-state.test.mjs +1640 -0
  48. package/scripts/ci-vacuous-green.test.mjs +368 -0
  49. package/scripts/citation-sweep-prose.test.mjs +1572 -0
  50. package/scripts/claim-lifecycle.test.mjs +185 -0
  51. package/scripts/claim-ticket.sh +417 -0
  52. package/scripts/claim-ticket.test.mjs +1121 -0
  53. package/scripts/closing-keyword-prose.test.mjs +193 -0
  54. package/scripts/cockpit-launch-idempotent-prose.test.mjs +165 -0
  55. package/scripts/cockpit-tick-nonblocking.test.mjs +210 -0
  56. package/scripts/compute-board.mjs +579 -0
  57. package/scripts/compute-board.test.mjs +1060 -0
  58. package/scripts/compute-spend.mjs +329 -0
  59. package/scripts/compute-spend.test.mjs +451 -0
  60. package/scripts/cross-repo-citation-prose.test.mjs +156 -0
  61. package/scripts/cwd-isolation-pins.mjs +83 -0
  62. package/scripts/dedupe-guard-prose.test.mjs +243 -0
  63. package/scripts/derive-testcmd.sh +239 -0
  64. package/scripts/derive-testcmd.test.mjs +380 -0
  65. package/scripts/diff-stats.mjs +231 -0
  66. package/scripts/diff-stats.test.mjs +408 -0
  67. package/scripts/dispatch-block-golden-prose.test.mjs +1024 -0
  68. package/scripts/dispatch-block-pins-prose.test.mjs +615 -0
  69. package/scripts/drop-merged-label.sh +173 -0
  70. package/scripts/drop-merged-label.test.mjs +480 -0
  71. package/scripts/expected-jobs-prose.test.mjs +80 -0
  72. package/scripts/filing-bar-seat-prose.test.mjs +242 -0
  73. package/scripts/filing-claim-bar-prose.test.mjs +166 -0
  74. package/scripts/finisher-dispatch-premise-prose.test.mjs +312 -0
  75. package/scripts/finisher-halt-prose.test.mjs +106 -0
  76. package/scripts/finisher-name-prose.test.mjs +134 -0
  77. package/scripts/finisher-own-tree-prose.test.mjs +285 -0
  78. package/scripts/finisher-pin-race-prose.test.mjs +339 -0
  79. package/scripts/fix-applier-correction-rules-prose.test.mjs +201 -0
  80. package/scripts/fix-applier-result-file-prose.test.mjs +100 -0
  81. package/scripts/fleet-bootstrap +155 -0
  82. package/scripts/fleet-heartbeat.mjs +272 -0
  83. package/scripts/fleet-heartbeat.test.mjs +450 -0
  84. package/scripts/fleet-provenance +207 -0
  85. package/scripts/fleet-provenance.test.mjs +64 -0
  86. package/scripts/fleet-run +148 -0
  87. package/scripts/fleet-run.test.mjs +90 -0
  88. package/scripts/fleet-state.mjs +400 -0
  89. package/scripts/fleet-state.test.mjs +373 -0
  90. package/scripts/fleet-tick-prose.test.mjs +299 -0
  91. package/scripts/fleet-tick.mjs +832 -0
  92. package/scripts/fleet-tick.test.mjs +1332 -0
  93. package/scripts/force-with-lease-reason-prose.test.mjs +184 -0
  94. package/scripts/frontmatter-allowlist.json +64 -0
  95. package/scripts/frontmatter-check.mjs +256 -0
  96. package/scripts/frontmatter-check.test.mjs +436 -0
  97. package/scripts/git-env.mjs +124 -0
  98. package/scripts/git-env.test.mjs +149 -0
  99. package/scripts/git-status-untracked-mode-sweep.test.mjs +216 -0
  100. package/scripts/gone-walk.test.mjs +213 -0
  101. package/scripts/gp-sep-invariant-prose.test.mjs +188 -0
  102. package/scripts/head-moved-after-label-prose.test.mjs +233 -0
  103. package/scripts/immutable-body-claim-prose.test.mjs +413 -0
  104. package/scripts/implementer-model-tier.test.mjs +643 -0
  105. package/scripts/inflight-citation-prose.test.mjs +144 -0
  106. package/scripts/inflight-exit2-prose.test.mjs +81 -0
  107. package/scripts/inflight-probe-guard-prose.test.mjs +109 -0
  108. package/scripts/inflight-scan-scope-prose.test.mjs +57 -0
  109. package/scripts/inflight.sh +1106 -0
  110. package/scripts/inflight.test.mjs +2984 -0
  111. package/scripts/injection-control-prose.test.mjs +404 -0
  112. package/scripts/install-root-audit.test.mjs +464 -0
  113. package/scripts/instrument-check-prose.test.mjs +235 -0
  114. package/scripts/instruments.sh +394 -0
  115. package/scripts/instruments.test.mjs +787 -0
  116. package/scripts/issue-tracker-prose.test.mjs +71 -0
  117. package/scripts/json.sh +248 -0
  118. package/scripts/json.test.mjs +364 -0
  119. package/scripts/ledger-dispatch.test.mjs +516 -0
  120. package/scripts/ledger-grammar.mjs +85 -0
  121. package/scripts/ledger-grammar.test.mjs +142 -0
  122. package/scripts/ledger-read-require-file.test.mjs +197 -0
  123. package/scripts/ledger.mjs +1753 -0
  124. package/scripts/ledger.test.mjs +3554 -0
  125. package/scripts/line-distance-prose.test.mjs +245 -0
  126. package/scripts/liveness-rationale-prose.test.mjs +83 -0
  127. package/scripts/locale-pin-prose.test.mjs +189 -0
  128. package/scripts/member-outcome-prose.test.mjs +167 -0
  129. package/scripts/member-outcomes-header.test.mjs +148 -0
  130. package/scripts/member-outcomes-prose.test.mjs +63 -0
  131. package/scripts/member-outcomes.mjs +231 -0
  132. package/scripts/member-outcomes.test.mjs +507 -0
  133. package/scripts/member-prompt-prose.test.mjs +161 -0
  134. package/scripts/member-record.mjs +605 -0
  135. package/scripts/member-record.test.mjs +516 -0
  136. package/scripts/member-write-guard.mjs +279 -0
  137. package/scripts/member-write-guard.test.mjs +211 -0
  138. package/scripts/merge-commit-probe-prose.test.mjs +122 -0
  139. package/scripts/merge-gate.mjs +262 -0
  140. package/scripts/merge-gate.test.mjs +613 -0
  141. package/scripts/muted-git-guard-sweep.test.mjs +493 -0
  142. package/scripts/net-ssh-precedence-prose.test.mjs +91 -0
  143. package/scripts/net-watchdog-prose.test.mjs +79 -0
  144. package/scripts/net.sh +409 -0
  145. package/scripts/net.test.mjs +211 -0
  146. package/scripts/no-ci-gate-prose.test.mjs +108 -0
  147. package/scripts/no-undo-audit.sh +1134 -0
  148. package/scripts/no-undo-audit.test.mjs +3965 -0
  149. package/scripts/no-wave.test.mjs +138 -0
  150. package/scripts/node-floor-sweep.test.mjs +611 -0
  151. package/scripts/pointer-target-prose.test.mjs +186 -0
  152. package/scripts/pr-overlap.mjs +417 -0
  153. package/scripts/pr-overlap.test.mjs +695 -0
  154. package/scripts/printf-die-sweep.test.mjs +384 -0
  155. package/scripts/prompt-renderer.mjs +39 -0
  156. package/scripts/prompt-renderer.test.mjs +68 -0
  157. package/scripts/prose-pin.mjs +808 -0
  158. package/scripts/prose-pin.test.mjs +1034 -0
  159. package/scripts/prove-merge.sh +228 -0
  160. package/scripts/prove-merge.test.mjs +995 -0
  161. package/scripts/pull-table-relabel-prose.test.mjs +60 -0
  162. package/scripts/quiet-payload-prose.test.mjs +222 -0
  163. package/scripts/reap.sh +1768 -0
  164. package/scripts/reap.test.mjs +4395 -0
  165. package/scripts/reaping-prose.test.mjs +159 -0
  166. package/scripts/refuter-scratch-prose.test.mjs +135 -0
  167. package/scripts/release-label-required-prose.test.mjs +71 -0
  168. package/scripts/release-label-split-prose.test.mjs +169 -0
  169. package/scripts/release-ticket.sh +1716 -0
  170. package/scripts/release-ticket.test.mjs +5106 -0
  171. package/scripts/renovate-release-contract.test.mjs +379 -0
  172. package/scripts/repo-root.mjs +544 -0
  173. package/scripts/repo-root.test.mjs +873 -0
  174. package/scripts/reporter-output-rule-prose.test.mjs +208 -0
  175. package/scripts/review-agent-report-only.test.mjs +49 -0
  176. package/scripts/review-core-cwd-isolation.test.mjs +374 -0
  177. package/scripts/review-core-reads.test.mjs +478 -0
  178. package/scripts/review-core-refuter-scratch.test.mjs +148 -0
  179. package/scripts/review-core-snapshot-path.test.mjs +777 -0
  180. package/scripts/review-core-specialist-scratch.test.mjs +189 -0
  181. package/scripts/review-core-testcmd.test.mjs +288 -0
  182. package/scripts/review-core-unrun.test.mjs +530 -0
  183. package/scripts/review-core-unverified-discriminant.test.mjs +208 -0
  184. package/scripts/review-core.mjs +1028 -0
  185. package/scripts/review-eval-resolver-path.test.mjs +55 -0
  186. package/scripts/review-eval.mjs +197 -0
  187. package/scripts/review-host-fixture.mjs +73 -0
  188. package/scripts/review-in-run-retry.test.mjs +187 -0
  189. package/scripts/review-no-vendor-dispatch.test.mjs +45 -0
  190. package/scripts/review-path-default.test.mjs +796 -0
  191. package/scripts/review-runner-spawns.test.mjs +93 -0
  192. package/scripts/review-runner.test.mjs +155 -0
  193. package/scripts/review-tier-audit.test.mjs +140 -0
  194. package/scripts/run-artifact-pr-prose.test.mjs +341 -0
  195. package/scripts/run-merge-bot-prose.test.mjs +1391 -0
  196. package/scripts/scratch-partition-invariant-prose.test.mjs +74 -0
  197. package/scripts/scripts-path-citation-sweep.test.mjs +73 -0
  198. package/scripts/section-between-guard-prose.test.mjs +378 -0
  199. package/scripts/select-dimensions.test.mjs +674 -0
  200. package/scripts/shared-refusal.test.mjs +388 -0
  201. package/scripts/shell-traps-prose.test.mjs +507 -0
  202. package/scripts/shortlist.mjs +355 -0
  203. package/scripts/shortlist.test.mjs +366 -0
  204. package/scripts/slow-transport.mjs +72 -0
  205. package/scripts/snapshot-repo.test.mjs +772 -0
  206. package/scripts/snapshot-runner-audience-prose.test.mjs +129 -0
  207. package/scripts/spend-view.test.mjs +312 -0
  208. package/scripts/staleness-qualifier-prose.test.mjs +251 -0
  209. package/scripts/staleness-verdict-prose.test.mjs +142 -0
  210. package/scripts/staleness.mjs +414 -0
  211. package/scripts/staleness.test.mjs +893 -0
  212. package/scripts/stalled-member-liveness-prose.test.mjs +219 -0
  213. package/scripts/stash-prohibition-prose.test.mjs +313 -0
  214. package/scripts/strip-comments.mjs +44 -0
  215. package/scripts/tier-check.mjs +468 -0
  216. package/scripts/tier-check.test.mjs +726 -0
  217. package/scripts/tier-guard-gate-prose.test.mjs +216 -0
  218. package/scripts/tier-outcomes-header.test.mjs +586 -0
  219. package/scripts/tier-outcomes.mjs +441 -0
  220. package/scripts/tier-outcomes.test.mjs +342 -0
  221. package/scripts/tier-roles.mjs +250 -0
  222. package/scripts/tier-roles.test.mjs +316 -0
  223. package/scripts/tracker-block-copy-prose.test.mjs +288 -0
  224. package/scripts/triage-labels-gloss-prose.test.mjs +97 -0
  225. package/scripts/unattended-git-sweep.test.mjs +142 -0
  226. package/scripts/verify-sha-exit2-prose.test.mjs +269 -0
  227. package/scripts/verify-sha.sh +117 -0
  228. package/scripts/verify-sha.test.mjs +693 -0
  229. package/scripts/watcher-degraded-block.test.mjs +372 -0
  230. package/scripts/watcher-key-prose.test.mjs +106 -0
  231. package/scripts/within-run-pair-prose.test.mjs +208 -0
  232. package/scripts/worktree-audit.sh +482 -0
  233. package/scripts/worktree-audit.test.mjs +1407 -0
  234. package/scripts/worktree-head-prose.test.mjs +276 -0
  235. package/scripts/worktree-listing-sweep.test.mjs +217 -0
  236. package/scripts/worktree-model-prose.test.mjs +216 -0
  237. package/scripts/worktree.sh +470 -0
  238. package/skills/next-ticket/SKILL.md +138 -0
  239. package/skills/run-team/SKILL.md +3603 -0
  240. package/skills/run-team/references/ci-and-staleness.md +75 -0
  241. package/skills/run-team/references/correction-tickets.md +37 -0
  242. package/skills/run-team/references/isolation.md +23 -0
  243. package/skills/run-team/references/member-lifecycle.md +73 -0
  244. package/skills/run-team/references/reaping.md +73 -0
  245. package/skills/sizing-a-ticket/SKILL.md +39 -0
@@ -0,0 +1,55 @@
1
+ import { test } from "node:test";
2
+ import assert from "node:assert/strict";
3
+ import { readFileSync, existsSync } from "node:fs";
4
+ import { join } from "node:path";
5
+
6
+ // Review finding on #1349's PR (#1361): `fleet-run scriptPathFor` joins
7
+ // `path.join(installPath, "scripts", script)` — the Resolver ALREADY places
8
+ // the caller under `scripts/`, and does no existence check on the result
9
+ // (`--path` "print the resolved path; do not exec"). Every prose site that
10
+ // told a caller to run `fleet-run --path scripts/review-eval.mjs` therefore
11
+ // resolved to `<installPath>/scripts/scripts/review-eval.mjs` — a path one
12
+ // level too deep that would only fail LOUDLY once something tried to
13
+ // `import()` it, never at the Resolver call itself. The correct invocation
14
+ // is `--path review-eval.mjs`, bare — `scripts/` is implicit.
15
+ //
16
+ // This is the existence check the Resolver itself does not perform: pin it
17
+ // here, against this repo's own `plugin/` tree standing in for an install
18
+ // root (`scriptPathFor`'s own join, replicated rather than imported — the
19
+ // Resolver is a standalone file placed outside the plugin by hand, per its
20
+ // own header, and importing it would need it to travel with a module it
21
+ // does not carry).
22
+ const REPO = join(import.meta.dirname, "..");
23
+
24
+ function scriptPathFor(installPath, script) {
25
+ return join(installPath, "scripts", script);
26
+ }
27
+
28
+ test("the correct --path form resolves to a file that actually exists", () => {
29
+ const correct = scriptPathFor(REPO, "review-eval.mjs");
30
+ assert.ok(existsSync(correct), `${correct} does not exist — review-eval.mjs moved or was renamed`);
31
+ });
32
+
33
+ test("the double-scripts/ form (the mistake this pins against) does not exist", () => {
34
+ const wrong = scriptPathFor(REPO, "scripts/review-eval.mjs");
35
+ assert.ok(
36
+ !existsSync(wrong),
37
+ `${wrong} exists — either a stray directory was created, or scriptPathFor's join changed and the prose sites need re-checking`,
38
+ );
39
+ });
40
+
41
+ // The prose sites: review-eval.mjs's own usage comment, plus every doc site
42
+ // that tells a caller how to load it. All three must spell the bare form.
43
+ const SITES = [
44
+ { path: ["scripts", "review-eval.mjs"], what: "review-eval.mjs's own usage comment" },
45
+ { path: ["commands", "review-and-fix.md"], what: "review-and-fix.md's OMP invocation line" },
46
+ { path: ["skills", "run-team", "SKILL.md"], what: "SKILL.md's OMP invocation line" },
47
+ ];
48
+
49
+ for (const { path, what } of SITES) {
50
+ test(`${what} spells the bare --path form, never the doubled scripts/ one`, () => {
51
+ const text = readFileSync(join(REPO, ...path), "utf8");
52
+ assert.match(text, /--path review-eval\.mjs\b/, `${what} no longer names the correct --path invocation`);
53
+ assert.doesNotMatch(text, /--path scripts\/review-eval\.mjs/, `${what} regressed to the doubled scripts/ form`);
54
+ });
55
+ }
@@ -0,0 +1,197 @@
1
+ // review-eval.mjs — the shim for the PR review port (#1349, per #1303's
2
+ // ruling on #1296). It is loaded (never review-core.mjs directly) through
3
+ // the Resolver, from an `eval` cell:
4
+ //
5
+ // const path = (await Bun.$`~/.fleet/bin/fleet-run --path review-eval.mjs`.text()).trim();
6
+ // const { runReviewOnOmp } = await import(path);
7
+ // const result = await runReviewOnOmp({ pr, branch, worktree, testCmd, scratch });
8
+ //
9
+ // Since #1802 the cell that does this is the `review-pr-<pr#>` member's own —
10
+ // agents/fleet-review-runner.agent.md, off the controller's turn — and it calls
11
+ // `runReviewToFile` (bottom of this file), which wraps `runReviewOnOmp` with the
12
+ // one retry and writes the result file. A controller holding its own turn can
13
+ // still call `runReviewOnOmp` exactly as above.
14
+ //
15
+ // This file `import`s review-core.mjs (a same-directory sibling, both ship
16
+ // together under the same Install root) with a RELATIVE specifier, so the
17
+ // Resolver is only needed ONCE, to find this file itself — the relationship
18
+ // between the two is an ordinary same-install sibling import, which eval's
19
+ // Bun VM permits without restriction (#1296 Q6). Never resolve
20
+ // review-core.mjs's own path through the Resolver a second time — that would
21
+ // be two doors where CONTEXT.md's Resolver entry says there is exactly one.
22
+ //
23
+ // No `model`/`effort` appears anywhere below (#1349 gap 3; audited by
24
+ // review-tier-audit.test.mjs). Every dispatch names a bare fleet-owned
25
+ // `agentType` from review-core.mjs's DEFAULT_DIMENSIONS or from
26
+ // runReview's own snapshot/verifier dispatch, unmodified — omp's `agent()`
27
+ // resolves a bare frontmatter `name:` exactly, which is already what those
28
+ // strings are.
29
+ import { mkdir, writeFile } from "node:fs/promises";
30
+ import { isAbsolute, join } from "node:path";
31
+ import { digestOf, runReview, runnerPrRefusal } from "./review-core.mjs";
32
+
33
+ // eval's `agent()` returns a HANDLE, not data (#1296 Q2): `agent(prompt,
34
+ // opts)` resolves near-instantly to an `AgentHandle` with `.wait()`, and only
35
+ // `.wait()` unwraps the schema-validated `structuredOutput.data`. This
36
+ // wrapper is what makes review-core.mjs's `await host.agent(...)` read like
37
+ // a synchronous, data-returning call — the ONE piece of host-specific glue
38
+ // `runReview` needs and cannot see.
39
+ //
40
+ // A crashed/rejected dispatch resolves to `null`, the contract every guard
41
+ // in review-core.mjs (`if (snap) {...}`, `unrunCrashed`, `verdictFor`) is
42
+ // written against. eval's DEFAULT `schemaMode` is "permissive": an exhausted
43
+ // structured-output retry is ACCEPTED anyway, carrying `schemaOverridden:
44
+ // true` on invalid data (#1296 Q2) rather than nulling out. `schemaMode:
45
+ // "strict"` turns that same exhaustion into a thrown/rejected result
46
+ // instead — still not a null — so this wrapper treats a REJECTED `.wait()`
47
+ // as the crash signal (`agent()`'s only way to fail loudly under omp) and
48
+ // maps it to `null` itself; it does not attempt to distinguish an
49
+ // accepted-but-schema-overridden permissive result from a clean one,
50
+ // because review-core.mjs never reads a distinguishing field for that case
51
+ // either — a schema violation the host already retried three times and
52
+ // gave up correcting is not this shim's contract to relitigate.
53
+ //
54
+ // NO PER-CALL WORKING DIRECTORY HERE, AND WHY THE ISOLATION RULE LIVES IN THE
55
+ // PROMPTS INSTEAD (#1433). Three PRs reviewed back-to-back from one eval cell
56
+ // left FOUR files modified in the checkout that cell was standing in — a
57
+ // specialist's mutation-test experiment reached through a RELATIVE path, in the
58
+ // tree the controller reads instruments.sh, ci-state.mjs and every gate
59
+ // decision out of. Review specialists have no assigned worktree to fall back
60
+ // to, so the preferred fix was a scratch cwd per dispatch, passed through this
61
+ // wrapper. The harness has no such option — measured against its own contract,
62
+ // not assumed:
63
+ //
64
+ // - eval's `agent()` takes `{ agent, label, schema, schemaMode, isolated,
65
+ // apply, merge, tools }` and nothing else (`omp://tools/eval.md`
66
+ // § `agent()`), and `task`'s item shape — `{ name, agent, task, effort,
67
+ // outputSchema, schemaMode, isolated }` — carries no cwd either
68
+ // (`omp://tools/task.md`). A non-isolated spawn "call[s] `runSubprocess(…)`
69
+ // directly with parent cwd", so the child's directory is decided one level
70
+ // above this file and is the controller's own.
71
+ // - `isolated: true` is not that option under another name. It names no
72
+ // directory, so it cannot BE the `<scratch>/pr<N>/<finding>/` the two-level
73
+ // scratch convention asks for; it exists only where
74
+ // `task.isolation.enabled` is on, and requesting it while isolation is
75
+ // `none` fails PREFLIGHT — synchronously, out of the `await agent(...)`
76
+ // below, which the `try` further down does not cover — so on any install
77
+ // with isolation off every dimension would come back null and the review
78
+ // would report itself wholly crashed. And it MERGES what the child changed
79
+ // back into the parent (patch apply, or a branch cherry-pick that stashes
80
+ // the parent repo first): a mutation-testing specialist's deliberately
81
+ // broken tree, applied to the checkout. That is the reported defect with a
82
+ // commit attached.
83
+ //
84
+ // So the rule lives where it can be stated at all: review-core.mjs's snapshot,
85
+ // specialist and refuter prompts name the inherited cwd as a no-run zone, order
86
+ // a `cd` into the run-root scratch path before any mutation, and require a
87
+ // `CWD-AUDIT:` line back. This wrapper stays a pure handle adapter. Should
88
+ // eval's `agent()` ever gain a cwd option, THIS is the one place that changes —
89
+ // thread it through `opts` here, and the prompt rules become the second belt
90
+ // rather than the only one.
91
+ async function ompAgent(prompt, opts) {
92
+ const handle = await agent(prompt, { agent: opts.agentType, schema: opts.schema, label: opts.label });
93
+ try {
94
+ return await handle.wait();
95
+ } catch {
96
+ return null;
97
+ }
98
+ }
99
+
100
+ // `pipeline()`/`parallel()` have no omp counterpart (#1296 Q3) — this is the
101
+ // same hand-rolled pair review-core.mjs exports as its own default (see
102
+ // `defaultPipeline`/`defaultParallel` there), passed through explicitly here
103
+ // rather than left to the default so a reader of THIS file, the one the
104
+ // controller actually loads, can see the omp orchestration shape without
105
+ // following an import.
106
+ function pipeline(items, stage1, stage2) {
107
+ return Promise.all(
108
+ items.map(async (item) => {
109
+ let r1;
110
+ try {
111
+ r1 = await stage1(item);
112
+ } catch {
113
+ r1 = null;
114
+ }
115
+ if (!r1) return null;
116
+ try {
117
+ return await stage2(r1, item);
118
+ } catch {
119
+ return null;
120
+ }
121
+ }),
122
+ );
123
+ }
124
+ function parallel(fns) {
125
+ return Promise.all(fns.map((fn) => fn()));
126
+ }
127
+
128
+ // The controller's `eval` cell entry point. `phase`/`log` are eval's own
129
+ // prelude globals (`omp://tools/eval.md`) — real free variables in the
130
+ // cell's scope, exactly like `agent` above; naming them as parameters here
131
+ // would ask the controller's cell to thread through globals it already has.
132
+ export async function runReviewOnOmp(args) {
133
+ return runReview({ agent: ompAgent, phase, log, pipeline, parallel }, args);
134
+ }
135
+
136
+ // #1802 (spec 2026-09-24-slot-based-fleet-loop-design.md § 3 §1, §2, §5, §7).
137
+ // The whole of the omp review runner's job, so the agent's own cell is three
138
+ // lines and this contract can be run (review-runner.test.mjs): the full result
139
+ // object goes to `<scratch>/review-<pr>.json` — the one artefact handed to
140
+ // the fix-applier — and only the digest comes back, because the digest is
141
+ // all the controller reads and a 26–61 KB result is what the file exists to
142
+ // keep out of its context.
143
+ //
144
+ // Failure is a throw or an empty return. The holder of the review call retries
145
+ // once; a second failure returns `failed` with both errors and the fallback
146
+ // reviewer's name (`review-pr-<pr>-b`), and writes no file, so nothing reads a
147
+ // half-review as a review. A dispatch mistake — a pr that is not a PR number, a
148
+ // scratch that is not absolute (eval's cwd is the MAIN CHECKOUT, so a relative
149
+ // one would put the file there) — throws before any run: it is not a review
150
+ // failure, and retrying or falling back would only repeat it. `run` is the seam
151
+ // the test injects; nothing else passes it.
152
+ export async function runReviewToFile(args, run = runReviewOnOmp) {
153
+ const pr = args?.pr;
154
+ const scratch = args?.scratch;
155
+ const prRefusal = runnerPrRefusal(pr);
156
+ if (prRefusal) throw new Error(`review-runner: ${prRefusal}`);
157
+ if (typeof scratch !== "string" || !isAbsolute(scratch)) {
158
+ throw new Error(`review-runner: args.scratch must be an absolute path, got ${JSON.stringify(scratch)}`);
159
+ }
160
+ const errors = [];
161
+ for (let attempt = 1; attempt <= 2; attempt++) {
162
+ let result;
163
+ try {
164
+ result = await run(args);
165
+ } catch (e) {
166
+ errors.push(`attempt ${attempt}: ${e?.message ?? String(e)}`);
167
+ continue;
168
+ }
169
+ // #1813: `result.counts` is read below to build the digest and the ledger
170
+ // token, but that read happens AFTER `mkdir`/`writeFile` have already put
171
+ // the file on disk — so a `run` that resolves to a truthy, object-shaped
172
+ // result missing `.counts` would otherwise throw on the destructure below
173
+ // OUTSIDE this function's own retry try/catch (which wraps only the
174
+ // `run(args)` call), leaving a half-written `review-<pr>.json` on disk
175
+ // that a fix-applier reading it would mistake for a completed review that
176
+ // never actually finished. Treated the same as an empty return: retried,
177
+ // then reported `failed`, and no file is written for it.
178
+ if (!result || typeof result !== "object" || !result.counts || typeof result.counts !== "object") {
179
+ errors.push(`attempt ${attempt}: empty return (${String(result)})`);
180
+ continue;
181
+ }
182
+ const path = join(scratch, `review-${pr}.json`);
183
+ await mkdir(scratch, { recursive: true });
184
+ await writeFile(path, `${JSON.stringify(result, null, 2)}\n`);
185
+ const { survived, refuted, unverified } = result.counts;
186
+ return {
187
+ status: "completed",
188
+ path,
189
+ // Spec § 3 §7's result token, ready for `ledger.mjs row`.
190
+ ledger: `reviewed=${result.head}:${survived}/${refuted}/${unverified}`,
191
+ attempts: attempt,
192
+ errors,
193
+ digest: digestOf(result),
194
+ };
195
+ }
196
+ return { status: "failed", errors, fallback: `review-pr-${pr}-b` };
197
+ }
@@ -0,0 +1,73 @@
1
+ // Shared `runReview(host, args)` test double, extracted from
2
+ // review-in-run-retry.test.mjs so review-path-default.test.mjs's own return-
3
+ // shape pin can drive a REAL `runReview` call instead of re-parsing
4
+ // review-core.mjs's source text for its `return {` — the retired script's
5
+ // own return literal was the ONLY thing that pin ever read;
6
+ // ADR 0014 (omp-only) leaves review-core.mjs as the one body to run.
7
+ //
8
+ // review-core.mjs's own pipeline contract, mirrored here: a null stage-1
9
+ // result never reaches stage 2, with a throw landing in the same null slot —
10
+ // the shape `defaultPipeline` (review-core.mjs) already has, restated as a
11
+ // scripted double so a caller needs no host beyond what it drives itself.
12
+ export async function pipeline(items, stage1, stage2) {
13
+ return Promise.all(
14
+ items.map(async (item) => {
15
+ let r1;
16
+ try {
17
+ r1 = await stage1(item);
18
+ } catch {
19
+ r1 = null;
20
+ }
21
+ if (!r1) return null;
22
+ try {
23
+ return await stage2(r1, item);
24
+ } catch {
25
+ return null;
26
+ }
27
+ }),
28
+ );
29
+ }
30
+ export const parallel = (fns) => Promise.all(fns.map((fn) => fn()));
31
+
32
+ export const ARGS = { pr: 7, branch: "feature/x", worktree: "/repo/.worktrees/7-x", scratch: "/scr", testCmd: "node --test", dimensions: ["correctness"] };
33
+ export const SNAP = {
34
+ runRoot: "/scr/pr7/run-ab12",
35
+ path: "/scr/pr7/run-ab12/snapshot-abc123",
36
+ head: "abc123",
37
+ pathVerified: true,
38
+ repoVerified: true,
39
+ testCmd: "node --test",
40
+ };
41
+ export const RAN = { command: "node --test", tests: 5, pass: 5, fail: 0 };
42
+ export const review = (findings, testRun = RAN) => ({
43
+ dimension: "correctness",
44
+ scope_searched: "CWD-AUDIT: clean /repo",
45
+ findings,
46
+ test_run: testRun,
47
+ });
48
+ export const finding = (severity) => ({ severity, claim: `a ${severity} claim`, file: "a.js", line: 3, evidence: "line 3 has no else" });
49
+ export const vote = (refuted) => ({ refuted, reason: "measured. CWD-AUDIT: clean /repo" });
50
+
51
+ // A host whose `agent()` answers each dispatch label from a script, one entry
52
+ // per call in dispatch order (the last entry repeats), and counts the calls.
53
+ // An `Error` entry is thrown rather than returned. Counting happens
54
+ // synchronously on the call, so the two refuters of one pair are calls 1 and
55
+ // 2 of their label, and a re-dispatched pair is calls 3 and 4.
56
+ export function scriptedHost(script) {
57
+ const calls = {};
58
+ return {
59
+ calls,
60
+ host: {
61
+ agent: async (_prompt, opts) => {
62
+ const n = (calls[opts.label] = (calls[opts.label] ?? 0) + 1);
63
+ const seq = script[opts.label];
64
+ if (!seq) throw new Error(`scriptedHost: unexpected dispatch ${opts.label}`);
65
+ const answer = seq[Math.min(n, seq.length) - 1];
66
+ if (answer instanceof Error) throw answer;
67
+ return answer === null ? null : structuredClone(answer);
68
+ },
69
+ phase: () => {},
70
+ log: () => {},
71
+ },
72
+ };
73
+ }
@@ -0,0 +1,187 @@
1
+ // #1802 (spec docs/specs/2026-09-24-slot-based-fleet-loop-design.md § 3 §2, §6).
2
+ // Two changes to the review's own result, in review-core.mjs's own body:
3
+ //
4
+ // 1. Crash repair inside the run. Each crashed specialist is re-dispatched
5
+ // once, and each refuter pair whose every vote died is re-dispatched once
6
+ // as a pair, before the result is assembled. What still fails lands in
7
+ // `dimensionsUnrun` / `unverified` as before, and `resume` now means
8
+ // "crashed again after the in-run retry".
9
+ // 2. Digest first. The small fields a controller acts on lead the returned
10
+ // object and the bulky finding arrays trail it, so a large review's
11
+ // digest is available without depending on (or risking truncation from)
12
+ // the finding arrays that follow it (before this, `resume` was LAST).
13
+ //
14
+ // Driven through a SCRIPTED HOST rather than a mocked runReview: `agent`,
15
+ // `pipeline`, `parallel`, `phase`, `log` are exactly the parameters
16
+ // runReview takes off `host`, so the retry and the digest ordering are
17
+ // exercised as review-core.mjs's own wiring runs them, never a re-derived
18
+ // copy of what that wiring is supposed to do. The fixtures below are shared
19
+ // with review-path-default.test.mjs's own return-shape pin, via
20
+ // review-host-fixture.mjs.
21
+ import { test } from "node:test";
22
+ import assert from "node:assert/strict";
23
+ import { runReview, DIGEST_KEYS, digestOf } from "./review-core.mjs";
24
+ import { pipeline, parallel, ARGS, SNAP, review, finding, vote, scriptedHost } from "./review-host-fixture.mjs";
25
+
26
+ const COPIES = [["review-core.mjs", (host, args) => runReview({ ...host, pipeline, parallel }, args)]];
27
+
28
+ for (const [name, run] of COPIES) {
29
+ test(`${name}: a specialist that crashes once is re-dispatched, and its review lands`, async () => {
30
+ const { host, calls } = scriptedHost({ snapshot: [SNAP], "review:correctness": [null, review([])] });
31
+ const result = await run(host, ARGS);
32
+ assert.equal(calls["review:correctness"], 2, "a crashed specialist was not re-dispatched in-run");
33
+ assert.deepEqual(result.dimensionsUnrun, [], "the re-dispatched specialist's review was not used");
34
+ assert.equal(result.cwdAudit.length, 1);
35
+ });
36
+
37
+ test(`${name}: a thrown specialist dispatch is a crash too, and is re-dispatched`, async () => {
38
+ const { host, calls } = scriptedHost({ snapshot: [SNAP], "review:correctness": [new Error("spend limit"), review([])] });
39
+ const result = await run(host, ARGS);
40
+ assert.equal(calls["review:correctness"], 2);
41
+ assert.deepEqual(result.dimensionsUnrun, []);
42
+ });
43
+
44
+ test(`${name}: a specialist that crashes twice is dispatched exactly twice, then named unrun`, async () => {
45
+ const { host, calls } = scriptedHost({ snapshot: [SNAP], "review:correctness": [null] });
46
+ const result = await run(host, ARGS);
47
+ assert.equal(calls["review:correctness"], 2, "the retry is ONE re-dispatch, not a loop");
48
+ assert.equal(result.dimensionsUnrun.length, 1);
49
+ assert.equal(result.dimensionsUnrun[0].dimension, "correctness");
50
+ assert.match(result.dimensionsUnrun[0].reason, /returned nothing/);
51
+ });
52
+
53
+ test(`${name}: a specialist that RETURNED but ran no suite is not a crash — never re-dispatched`, async () => {
54
+ const { host, calls } = scriptedHost({
55
+ snapshot: [SNAP],
56
+ "review:correctness": [review([], { command: "node --test", tests: 0, pass: 0, fail: 0 })],
57
+ });
58
+ const result = await run(host, ARGS);
59
+ assert.equal(calls["review:correctness"], 1, "a returned-but-unrun review was re-dispatched as if it had crashed");
60
+ assert.equal(result.dimensionsUnrun.length, 1);
61
+ assert.match(result.dimensionsUnrun[0].reason, /0 tests/);
62
+ });
63
+
64
+ test(`${name}: a refuter pair whose every vote died is re-dispatched once, as a pair`, async () => {
65
+ const { host, calls } = scriptedHost({
66
+ snapshot: [SNAP],
67
+ "review:correctness": [review([finding("critical")])],
68
+ "verify:correctness": [null, null, vote(false), vote(false)],
69
+ });
70
+ const result = await run(host, ARGS);
71
+ assert.equal(calls["verify:correctness"], 4, "a wholly crashed refuter pair was not re-dispatched as a pair");
72
+ assert.equal(result.survived.length, 1);
73
+ assert.equal(result.survived[0].refutersDispatched, 2);
74
+ assert.deepEqual(result.counts, { survived: 1, refuted: 0, unverified: 0, crashed: 0 });
75
+ assert.equal(result.resume, null, "a pair the retry recovered still reads as crashed");
76
+ });
77
+
78
+ test(`${name}: a pair that crashes again after the retry is unverified, counted crashed, and resume says so`, async () => {
79
+ const { host, calls } = scriptedHost({
80
+ snapshot: [SNAP],
81
+ "review:correctness": [review([finding("critical")])],
82
+ "verify:correctness": [null],
83
+ });
84
+ const result = await run(host, ARGS);
85
+ assert.equal(calls["verify:correctness"], 4, "the pair retry is ONE re-dispatch, not a loop");
86
+ assert.equal(result.unverified.length, 1);
87
+ assert.equal(result.unverified[0].refutersDispatched, 2);
88
+ assert.deepEqual(result.counts, { survived: 0, refuted: 0, unverified: 1, crashed: 1 });
89
+ assert.match(result.resume, /in-run retry/, "resume no longer says the crash survived the in-run retry");
90
+ });
91
+
92
+ // #1813: a REJECTED dispatch (not a resolved `null`) is the shape every
93
+ // real `parallel()`/`agent()` failure actually takes on omp — see
94
+ // `retryCrashed`'s own doc comment: "that second answer is final, THROWN or
95
+ // not". Before this fix, a finding whose refuter pair rejected on both
96
+ // `retryCrashed` attempts propagated that rejection into the shared,
97
+ // findings-level `parallel()` (a bare `Promise.all`), which discarded every
98
+ // OTHER finding under the same dimension — including ones whose refuters
99
+ // fully succeeded — and reported the whole dimension `dimensionsUnrun`
100
+ // with the misleading "the reviewer returned nothing" reason, even though
101
+ // the specialist's review DID come back with real findings.
102
+ test(`${name}: a finding whose refuter pair rejects on both retry attempts does not erase its dimension-mates' verdicts`, async () => {
103
+ const crashing = finding("critical");
104
+ const surviving = finding("important");
105
+ const calls = { crashing: 0, surviving: 0 };
106
+ const host = {
107
+ agent: async (prompt, opts) => {
108
+ if (opts.label === "snapshot") return structuredClone(SNAP);
109
+ if (opts.label === "review:correctness") return structuredClone(review([crashing, surviving]));
110
+ if (opts.label === "verify:correctness") {
111
+ if (prompt.includes(crashing.claim)) {
112
+ calls.crashing++;
113
+ throw new Error(`refuter crash ${calls.crashing}`);
114
+ }
115
+ if (prompt.includes(surviving.claim)) {
116
+ calls.surviving++;
117
+ return structuredClone(vote(false));
118
+ }
119
+ throw new Error("scriptedHost: unexpected verify prompt");
120
+ }
121
+ throw new Error(`scriptedHost: unexpected dispatch ${opts.label}`);
122
+ },
123
+ phase: () => {},
124
+ log: () => {},
125
+ };
126
+ const result = await run(host, ARGS);
127
+ assert.equal(calls.crashing, 4, "the crashing pair is dispatched twice per attempt, across both retryCrashed attempts");
128
+ assert.equal(calls.surviving, 2, "the surviving pair's own refuters are undisturbed by the other finding's crash");
129
+ assert.equal(result.survived.length, 1, "a fully-verified sibling finding must not be erased by another finding's crash");
130
+ assert.equal(result.survived[0].claim, surviving.claim);
131
+ assert.equal(result.unverified.length, 1, "the crashed finding is reported unverified, not silently dropped");
132
+ assert.equal(result.unverified[0].claim, crashing.claim);
133
+ assert.equal(result.unverified[0].refutersDispatched, 2);
134
+ assert.deepEqual(result.dimensionsUnrun, [], "the dimension itself ran and returned real findings — it must not read as unrun");
135
+ assert.deepEqual(result.counts, { survived: 1, refuted: 0, unverified: 1, crashed: 1 });
136
+ });
137
+
138
+ test(`${name}: a pair with one live vote did not crash — it is ruled on that vote, not re-dispatched`, async () => {
139
+ const { host, calls } = scriptedHost({
140
+ snapshot: [SNAP],
141
+ "review:correctness": [review([finding("important")])],
142
+ "verify:correctness": [null, vote(true)],
143
+ });
144
+ const result = await run(host, ARGS);
145
+ assert.equal(calls["verify:correctness"], 2, "a pair with a surviving vote was re-dispatched as if every refuter had died");
146
+ assert.equal(result.refuted.length, 1);
147
+ assert.deepEqual(result.counts, { survived: 0, refuted: 1, unverified: 0, crashed: 0 });
148
+ });
149
+
150
+ test(`${name}: a suggestion's 0-refuter budget is policy, not a crash — nothing is dispatched for it`, async () => {
151
+ const { host, calls } = scriptedHost({ snapshot: [SNAP], "review:correctness": [review([finding("suggestion")])] });
152
+ const result = await run(host, ARGS);
153
+ assert.equal(calls["verify:correctness"], undefined, "a 0-refuter suggestion bought a refuter dispatch");
154
+ assert.equal(result.unverified.length, 1);
155
+ assert.equal(result.unverified[0].refutersDispatched, 0);
156
+ assert.deepEqual(result.counts, { survived: 0, refuted: 0, unverified: 1, crashed: 0 });
157
+ assert.equal(result.resume, null);
158
+ });
159
+
160
+ test(`${name}: the digest leads the returned object, so it is a prefix of the serialized result`, async () => {
161
+ const { host } = scriptedHost({
162
+ snapshot: [SNAP],
163
+ "review:correctness": [review([finding("critical"), finding("suggestion")])],
164
+ "verify:correctness": [vote(false)],
165
+ });
166
+ const result = await run(host, ARGS);
167
+ assert.deepEqual(Object.keys(result), [...DIGEST_KEYS, "snapshot", "survived", "refuted", "unverified"]);
168
+ // The consumer-visible property, not the key list: the pre-cutover
169
+ // harness hands the controller the FIRST ~8 KB of the serialized result,
170
+ // so every digest field has to be serialized before the first byte of a
171
+ // finding array.
172
+ const whole = JSON.stringify(result);
173
+ const digest = JSON.stringify(digestOf(result));
174
+ assert.ok(whole.startsWith(digest.slice(0, -1) + ","), "a bulk field is serialized ahead of the digest — a large review truncates it away");
175
+ });
176
+ }
177
+
178
+ // The accept half of the digest helper: every digest key, nothing else, and in
179
+ // the result's own order — a digest that dropped `resume` is the defect this
180
+ // reorder exists to fix, one layer up.
181
+ test("digestOf keeps exactly the digest keys, in order, and leaves the finding arrays behind", async () => {
182
+ const { host } = scriptedHost({ snapshot: [SNAP], "review:correctness": [review([finding("suggestion")])] });
183
+ const result = await runReview({ ...host, pipeline, parallel }, ARGS);
184
+ const digest = digestOf(result);
185
+ assert.deepEqual(Object.keys(digest), DIGEST_KEYS);
186
+ for (const bulk of ["snapshot", "survived", "refuted", "unverified"]) assert.equal(bulk in digest, false, bulk);
187
+ });
@@ -0,0 +1,45 @@
1
+ import { test } from "node:test";
2
+ import assert from "node:assert/strict";
3
+ import { readFileSync, readdirSync } from "node:fs";
4
+ import { join } from "node:path";
5
+
6
+ // #1349's own acceptance text: `grep -n 'pr-review-toolkit' skills/
7
+ // commands/` returns nothing — the six vendored names are never DISPATCHED.
8
+ // Scoped to those two directories deliberately, not the whole `plugin/`
9
+ // tree: `agents/fleet-review-*.agent.md`'s credit comments plainly name the
10
+ // vendored source they adapted a prompt from (review finding on #1361 — an
11
+ // earlier revision obfuscated the slug there to satisfy an over-widened
12
+ // reading of this same criterion, which helped nobody and made the credit
13
+ // harder to verify). Attribution prose and dispatch names are different
14
+ // claims; this pins the one #1349 actually makes.
15
+ const REPO = join(import.meta.dirname, "..");
16
+
17
+ function allFiles(dir) {
18
+ const out = [];
19
+ for (const entry of readdirSync(dir, { withFileTypes: true })) {
20
+ const full = join(dir, entry.name);
21
+ if (entry.isDirectory()) out.push(...allFiles(full));
22
+ else out.push(full);
23
+ }
24
+ return out;
25
+ }
26
+
27
+ test("no file under skills/ or commands/ mentions the retired vendor plugin", () => {
28
+ for (const dir of ["skills", "commands"]) {
29
+ for (const file of allFiles(join(REPO, dir))) {
30
+ const text = readFileSync(file, "utf8");
31
+ assert.doesNotMatch(text, /pr-review-toolkit/, `${file} still mentions pr-review-toolkit`);
32
+ }
33
+ }
34
+ });
35
+
36
+ // The counterpart: agents/ deliberately DOES name it, in plain text, as
37
+ // attribution — this is not covered by the criterion above and must not be
38
+ // "fixed" into the same obfuscated shape a prior revision tried.
39
+ test("agents/fleet-review-*.agent.md credit the vendored source plainly, not obfuscated", () => {
40
+ const dims = ["correctness", "silent-failure", "tests", "comments", "types", "simplify"];
41
+ for (const key of dims) {
42
+ const text = readFileSync(join(REPO, "agents", `fleet-review-${key}.agent.md`), "utf8");
43
+ assert.match(text, /pr-review-toolkit/, `fleet-review-${key}.agent.md no longer plainly credits the vendored plugin`);
44
+ }
45
+ });