Repository navigation
review: spec-first fanout and ratchet move to toolchain #7584
Description
Parent
Refs #7555. Source: PLAN Phase 16, topic poteto-review. Thread: review.
Depends on
- config-cascade: settings resolution and convention home #7571 Settings resolution and convention home
- standards: per-concern standards layout #7583 Standards layout (per-concern files feed fanout criteria slices)
Outcome
/review:fanout reports spec findings first under their own heading, then standards, with an act-on
cap set from an eval; /review:ratchet becomes /toolchain:ratchet with a per-file baseline mode.
Decisions
/review:fanoutgains Lauren's interrogate lead judgment: spec findings first under their own
heading, then standards; act-on cap per heading (setting from the Q160 eval); agreement map covers
both; stays a code review (D101, Q159). Origin: combined (Lauren's interrogate; Matt keeps the
two axes separate).- Criteria from per-concern standards files loaded as criteria slices; a code-quality lens slice with
a configurable file-size threshold that only flags long files on targeted work (D101, Q135). - Fanout documented for design stub scaffolds (Q169). Review criteria stay with the reviewer;
implementation loads only paved-path entries for the paths it touches (D98, Q153). - Standards discovery names a root
CODING_STANDARDS.mdandCONTRIBUTING.mdas rung-4 sources
(Q523). /review:ratchetmoves to/toolchain:ratchet; name stays; the limit is "the ceiling"; per-file
baseline mode using each linter's native baseline (D36, D63, D81, Q513; Lauren's "fail only
when a change adds more"). The zero-entry baseline rule (Q181) lives in the/toolchain:ratchet
reference (B8).- Lint rollout: codemod then error when mechanical, else per-file baseline at error severity; break
glass is an inline exception with reason, expiry and a human's approval, as a team setting (D64,
Q23). - Research tag: fanout reviewer-diversity eval (12 diffs, 4 arms, 3 runs) settles the lens default,
panel placement and the act-on cap (Q134, Q160, D101).
Key interfaces
From the design's module layout: "review: /review:fanout gains spec-first judgment and a
code-quality lens slice; /review:audit-enforceability moves to corrections; /review:ratchet
moves to /toolchain:ratchet." "toolchain: gains /toolchain:ratchet (per-file baseline mode)."
Settings: "fanout act-on cap from the Q160 eval."
Slices, filed when this phase starts: fanout-spec-first, fanout-eval-and-cap, ratchet-move.
Done when
- The ratchet skill lives under toolchain, not review, and
rg -n 'review:ratchet' plugins docs --glob '!**/CHANGELOG.md' --glob '!.changes/**'prints
nothing. - The five zero-entry baseline files that exist on main are deleted with their reading code.
- The fanout eval slice carries a
Probe result:comment before the act-on cap merges.
Out of scope
/review:audit-enforceability(moves with the corrections thread).
Activity
github-actions commentedon Oct 16, 2026
Proposed work class: work-class: structural
Basis: automated intake triage; the lanes App applied this class.
Size: Large
Complexity: Medium
Uncertainty: Knowable
Flags: none
Blocked by: none
Understood: this moves
/review:ratchetto/toolchain:ratchetwith a per-file baseline mode. It also makes/review:fanoutreport spec findings first under their own heading, then standards. The act-on cap would come from the reviewer-diversity eval. The issue also lists the five zero-entry baseline files to delete.The work covers three slices (fanout-spec-first, fanout-eval-and-cap, ratchet-move), so it needs splitting into sub-issues. The decisions are already written in the body, so I'm asking no new questions. #7571 and #7583 are not in the open-issue list I was given, so I could not confirm whether they are still open. I found no likely duplicates.
Automated triage, applied by the lanes App.
kyle-sexton commentedon Oct 24, 2026
Released for re-triage under the owner's 2026-10-24 rule: an issue stops for a person only for an intent or vision question, a security call, a host or attended run, or spend outside prepaid credit. The hold label came off and intake triage runs again now; it re-applies a hold, with a numbered question, when one of those applies.
To hold this issue again, add
needs-human(orstatus: needs-decision) and say why in a comment.13 remaining items
kyle-sexton commentedon Nov 2, 2026
Status: what landed and what is left.
Status check by the issue-triage session against main at
0ed6022913.Met:
rg -n 'review:ratchet' plugins docs --glob '!**/CHANGELOG.md' --glob '!.changes/**'prints nothing:plugins/toolchain/skills/ratchet/exists andplugins/review/skills/ratchet/does not. Moved by refactor(toolchain): move the ratchet skill from review to toolchain #11355 (a7bca17ac8).e6930abcda):scripts/changelog-parity-baseline.txt,scripts/hook-userconfig-argv-allowlist.txtandscripts/skill-description-cap-baseline.txt, with their reads incheck-changelog-parity.sh,check-hook-userconfig-argv.shandcheck-changed-skills.sh.f11db29358):scripts/test-tmp-cleanup-baseline.txtandscripts/verb-contract-baseline.txt, with their reads incheck-test-tmp-cleanup.shandcheck-all-skills-verb-contract.sh.Remaining:
Probe result:comment before the act-on cap merges. The eval slice is test(review): run the fanout reviewer-diversity eval and post its probe result #12151, still open; no act-on cap has merged (the phrase appears only inplugins/review/evals/fanout-diversity.md). The probe needs aclaude plugin evalrun of the fanout reviewer-diversity suite, which is a host run.test(review): lead the fanout-diversity verdict with its marker and s…
kyle-sexton commentedon Nov 3, 2026
Rule (3) measured, lower rigor (Codex arms outside the harness, on the host's own Codex login): L+X recall 23/23 against L 24/24, difference 0 at ceiling, so
/codex:reviewstays out of the default sweep. Details: #12151 (comment). #12505's harder fixtures are what can separate the arms.