Repository navigation
fix(source-control): require the legacy discovered stamp in publish, and test the re-earn exclusion #12450
Description
Problem
The independent security review of #12394's legacy discovery stamp (commit ee07d83, verdict PASS) left two suggestions in plugins/source-control/skills/babysit-loop/scripts/publish-promotion-evidence.mjs:
publish()(~L1000-1004) allows the one-time legacy stamp but doesn't require it. A prior contrary event that stays withoutdiscovered, or a new contrary event without one, passes (undefined === undefined); enforcement rests oncollect(). The only path throughmain()is a race: the sidecar is absent or unparsable whenmainreads the out directory, then reappears beforepublish()re-reads it.collectthen derives a re-earn withsince = at. That isn't reachable by anyone who can't write the out directory, and the commit before feat(autonomy): re-promote a demoted cell once its predicate is met again #12394 behaved the same way.- No test covers the
event?.event !== "re-earned"clause, i.e. that a prior re-earned event gaininglegacyDiscovered(previousAt)is refused.
Fix
- In
--outmode,publish()refuses any non-re-earned event whosediscoveredis missing or doesn't parse. The daily-loop test's legacy fixture is then written directly, not built throughpublish(). - Add
assert.throws(() => publish(out, [{...R, discovered: legacyDiscovered}]))over a file holding a re-earnR.
Done when
- Both changes land with tests, and the existing publisher (82) and end-to-end (36) suites still pass.
Activity
github-actions commentedon Nov 2, 2026
Proposed work class: work-class: scoped
Basis: automated intake triage; the lanes App applied this class.
Size: Small
Complexity: Low
Uncertainty: Known
Flags: none
Blocked by: none
#12450 (comment)
I understood this as two changes. In
--outmode,publish()should refuse any non-re-earned event whosediscoveredstamp is missing or unparsable. A new test should cover a prior re-earned event gaining the legacy stamp. The recorded decision settles the open questions, so nothing further is being asked.Automated triage, applied by the lanes App.
melodic-automation-lanes commentedon Nov 2, 2026
Not admitted to the issue lane: it carries host-run. Admission takes the issue once each of those is met, when intake triages it again (
.github/actions/resolve-config/admission-rules.json).melodic-automation-lanes commentedon Nov 2, 2026
The latest triage verdict holds nothing: admission moved the issue to
status: confirmed; its comment says why.melodic-automation-lanes commentedon Nov 2, 2026
Refinement. A read-only run explored this repository, searched its issues and pull requests, and read the allowed docs to draft what follows. On an issue a listed human filed, it counts as the issue's spec once posted, while unedited; it is still model text, so check it.
Intended outcome: In --out mode publish() refuses any non-re-earned event whose discovered stamp is missing or unparsable, and a test shows a prior re-earn gaining legacyDiscovered is refused.
Related work
Open questions
discoveredorlegacyDiscoveredanywhere in the babysit-loop skill, and publish() at publish-promotion-evidence.mjs:897 has no stamp logic. Has the #12394 stamp (commit ee07d83) landed on the base branch, or should this issue wait for it?RECOMMENDED: Treat #12394's stamp change as a prerequisite and hold this issue until it is on the base branch. Basis: plugins/source-control/skills/babysit-loop/scripts/publish-promotion-evidence.mjs:897
RECOMMENDED: Name the daily-loop test file and the commands that run the two suites in the issue, so a run can check the counts. Basis: judgment
Sources read
To answer, post a comment that starts with this line:
Automated refinement by the lanes App. Answer with a comment whose first line is
**Decision recorded.**.kyle-sexton commentedon Nov 3, 2026
Decision recorded.
The #12394 stamp is on main (
publish-promotion-evidence.mjs:290,:1002at origin/main), so nothing blocks this fix. The publisher suite isplugins/source-control/skills/babysit-loop/scripts/publish-promotion-evidence.test.mjs(run bypublish-promotion-evidence.test.sh); the body does not name the file holding the 36 end-to-end tests, so the builder states it in the PR. Written by the issue-triage session.--outmode,publish()refuses any non-re-earned event whosediscoveredis missing or doesn't parse. The daily-loop test's legacy fixture is then written directly, not built throughpublish().assert.throws(() => publish(out, [{...R, discovered: legacyDiscovered}]))over a file holding a re-earnR.kyle-sexton commentedon Nov 3, 2026
Lane 3 claims this for the overnight host-run pass.
kyle-sexton commentedon Nov 3, 2026
Agent choice, not an owner answer. The decision record above was written by the issue-triage session under the owner's standing direction §1 (decide from evidence). The owner did not answer this question; the comment appears under the owner's login only because the session posts with that account.
github-actions commentedon Nov 3, 2026
The triage sweep dispatched intake triage for this issue (attempt 1 of 2).