ci(claude): let a rewritten comment drive the reconcile round

[Troublor]

Problem

megaeth-labs/mega-agents#317 is the demonstration: mega-putin answered mega-maxwell's open question and the reviewer never noticed.

time	event
12:15:12	mega-putin posts :mag: Looking into this... (26 chars)
12:15:15	issue_comment.created → review round runs. Run log: conversation: 1 of 1 timeline entries · 26 chars — it reconciled the placeholder
12:16:11	status comment republished: still "1 open question"
12:20:31	mega-putin edits that same comment into the real answer → no event, no round, ever

Not a bot-filtering problem — allowed_bots: mega-putin is set and the run log shows the round proceeding under mega-putin[bot]. The round simply landed on the wrong revision of the comment.

Change

• issue_comment: types: [created, edited], so the rewrite that carries the answer drives the reconcile round.
• pr-review declines comments carrying claude-review:skip — the marker mega-putin now puts on a placeholder it intends to rewrite (megaeth-labs/mega-agents#318). The shared action enforces the same marker (fix(pr-review): let a rewritten comment drive the reconcile round .github#46, merged); the job if declines one step earlier, before a runner starts, so the placeholder costs nothing.
• interactive pins github.event.action == 'created', so editing an old comment that mentions @claude cannot re-run a session.

The mega-maxwell[bot] exclusion already present is what keeps the sticky status comment — rewritten in place every round — from retriggering under edited.

Same change is going into every repo that opted into comment-triggered reconciliation. dist-docs is excluded: its pr-review never subscribed to issue_comment.

Test plan

Workflow-only. YAML parses; the if conditions are the existing ones plus one conjunct each.

🤖 Generated with Claude Code

https://claude.ai/code/session_01DMw1H8WTihb5CfQ7tabxdt

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-14T12:43:29.881515Z	07a4af1	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[mega-maxwell]

Claude review status

Living comment — rewritten in place. The review workflow keeps this single comment up to date instead of posting a new one each round, so it always describes the latest reviewed commit and the earlier text is intentionally gone. No reply is needed here; reply to a finding in its own review thread, and answer an open question in a reply on this PR. The next review round reconciles your answer.

🛠️ Review did not finish

Attempted head 07a4af14 · updated 2026-09-14T12:41:27+00:00

This round did not publish: MODEL_NO_OUTPUT in phase compile. Anything listed below is from the last round that did. Re-run the workflow or push a new commit to try again.