Repository navigation
maintainer-resolve-comment.sh @-mentions the bot it is clearing, so Codex posts a new blocking comment every time #2255
Description
Problem
scripts/maintainer-resolve-comment.sh clears a registered bot's status comment (Codex usage notice, SonarCloud clean status, and so on) by posting a reply and then minimizing the original. The reply begins Maintainer disposition of @<bot>'s status comment by @<maintainer>: (line 223). That @<bot> is a live mention.
Codex answers every mention. So clearing a Codex usage-limit notice makes Codex post a new comment a few seconds later, and the PR is blocked by the new comment instead of the old one. The tool cannot win against Codex.
Evidence
- PR feat: implement issue #1900 — [Phase 3] Refresh the Opus 4.8 fallback-chain hops to Opus 5.5 in engine.sh #2241, 2026-10-11: the tool resolved the Codex notice at 02:28:07Z; Codex posted "You have reached your Codex usage limits." again at 02:28:16Z.
- PR feat: implement issue #2079 — dev-lead: resolve a bot review thread on a maintainer "no change needed" disposition (rebuild of #1799 on #2049's resolver) #2084, 2026-10-09: the tool resolved a Codex notice at 23:42:04Z; Codex posted "To use Codex here, create an environment for this repo" at 23:42:14Z. That text does not match Codex's registered status pattern, so the tool could not clear it at all.
Acceptance criteria
- The reply the tool posts does not mention the bot or the maintainer. Name them without the
@(for example in backticks), keeping themaintainer-resolvemarker unchanged. - A bats test asserts the reply body contains no
@<login>for either account. - Codex's "To use Codex here, create an environment for this repo" notice is added to its
info_status_patterninscripts/lib/reviewer-sources.tsv, with a test, so an existing one can be cleared.
Activity
don-petry commentedon Oct 11, 2026
Dev-Lead Implementation Plan
Issue: #2255 — maintainer-resolve-comment.sh @-mentions the bot it is clearing, so Codex posts a new blocking comment every time
Scope
Stop the maintainer-resolve reply from @-mentioning the bot or the maintainer, and register Codex's "create an environment" notice as an info-status body so it can be cleared.
Implementation Checklist
scripts/maintainer-resolve-comment.sh—mrc_build_reply_bodynames the bot and the maintainer in backticks instead of@login; themaintainer-resolvemarker stays the samescripts/lib/reviewer-sources.tsv— addTo use Codex here, [create an environment for this repo](…).to thechatgpt-codex-connectorinfo_status_pattern, anchored at both ends like the existing usage-limit formsTests to Write
tests/dev-lead/unit/test_maintainer_resolve_comment.bats— the reply body contains no@<bot>and no@<maintainer>, still has the marker unchanged, and still names both accountstests/test_reviewer_sources.bats— the real environment notice from PR feat: implement issue #2079 — dev-lead: resolve a bot review thread on a maintainer "no change needed" disposition (rebuild of #1799 on #2049's resolver) #2084 matches Codex's pattern; quoted or finding-followed variants do notTest Command
bats tests/dev-lead/unit/test_maintainer_resolve_comment.bats tests/test_reviewer_sources.bats(plus the fullbatssuite)don-petry commentedon Oct 11, 2026
Dev-Lead: Progress (in progress — not yet durable)
Plan Execution (intended)
mrc_build_reply_bodynow names the bot and the maintainer in backticks instead of@login. Themaintainer-resolvemarker line is unchanged.info_status_patternnow also acceptsTo use Codex here, [create an environment for this repo](…)., with the link or without it. It is still anchored at both ends. The text was copied from the real comment on PR feat: implement issue #2079 — dev-lead: resolve a bot review thread on a maintainer "no change needed" disposition (rebuild of #1799 on #2049's resolver) #2084.Files touched (this workspace, not yet pushed)
scripts/maintainer-resolve-comment.sh: reply body with no @-mentionsscripts/lib/reviewer-sources.tsv: Codex pattern extended to the environment notice; the rationale column now cites maintainer-resolve-comment.sh @-mentions the bot it is clearing, so Codex posts a new blocking comment every time #2255tests/dev-lead/unit/test_maintainer_resolve_comment.bats: the reply contains neither@<bot>nor@<maintainer>, still names both, and keeps the markertests/test_reviewer_sources.bats: the environment notice clears; a quoted copy and a notice followed by a finding do notNotes
test_engine_fallback.batsandtest_engine_gemini_35.batsfail locally both with and without this change, so those failures already existed here and are not caused by it.don-petry commentedon Oct 11, 2026
Dev-Lead: Implementation Complete — PR #2265
The implementation for issue #2255 is durable: commits are pushed and a pull request is open. This record is backed by verifiable artifacts you can confirm in one click.
3b539dbe2754120d99c46beb438daea0ba749880main: 4 file(s), +34/-3 linesReview happens on the PR. Acceptance criteria and test results are evidenced by the PR's diff and CI checks — not asserted here. If a later stage fails, this claim is retracted in place.