Skip to content

fix(intake-triage): an issue refused for a body edited after deferral is never triaged again #18049

Open

Description

@kyle-sexton

Steps to reproduce

  1. Intake defers an issue (admission-deferred added).
  2. Someone edits the issue body.
  3. The next admission refuses it: "its body was edited after it was deferred, so the admission it was deferred from judged an older body. Admission takes the issue ... when intake triages it again" (.github/scripts/intake-triage/admit.mjs:1316).

Seen on 2026-11-21: 15 config-cascade issues (#17999-#18014, less #18008) refused this way at 19:15Z-19:23Z, with no label left on them, and still un-admitted at 21:17Z (metrics L982). The hub re-triaged them by hand with gh workflow run intake-triage.yml -f issue-number=<N>.

Expected behavior

A body edit on an issue refused for a stale body starts a fresh triage once, so the issue reaches the lanes without a person.

Actual behavior

No event re-triages it. intake-triage.yml triggers on issues: [opened, reopened] and workflow_dispatch only (lines 38-44), and intake-reenter.yml on issue_comment: created and issues: [unlabeled, field_added, field_removed] (lines 53-58). The edit that caused the refusal has already happened, so the issue sits until someone comments, relabels or dispatches.

Severity

Medium.

Severity justification

It blocks issue flow with no signal to a person: 15 issues sat un-admitted for 2 h today, and the refusal text promises a re-triage that nothing starts.

Suggested fix location

The smallest fix is one of:

  • in admit.mjs, when the only refusal is the stale-body one, re-triage at once against the current body (the edit is already in), instead of refusing; or
  • add edited to intake-reenter.yml's issues types, gated in its if: to issues whose newest intake comment is the stale-body refusal, so ordinary edits start nothing.

Prefer the first: it needs no new event and no call per edit.

Environment

claude-code-plugins main, 2026-11-21.

Activity

  1. added
    needs-triageNot yet classified. Floor until a type and one priority tier are set.
    on Nov 21, 2026
  2. github-actions commented on Nov 21, 2026

    @github-actions
    Contributor

    Size: Small
    Complexity: Low
    Uncertainty: Known
    Flags: none
    Blocked by: none
    Blast radius: intake-triage admission (admit.mjs, the refusal text in admission-comment.mjs, admit.test.mjs); it changes what happens to deferred issues whose body was edited.
    Reversible: yes; revert the change.
    Open questions: none
    Understood: an issue deferred and then edited is refused for a stale body, and nothing re-triages it. The issue picks the admit.mjs fix (send it back to intake) over a new edited trigger, and its acceptance criteria are exact. No duplicate open issue and no merged pull request that already fixes this turned up.

    Automated triage. A maintainer reviews it.

  3. removed
    needs-triageNot yet classified. Floor until a type and one priority tier are set.
    on Nov 21, 2026
  4. melodic-automation-lanes commented on Nov 21, 2026

    @melodic-automation-lanes
    Contributor

    Not admitted to the issue lane: no trusted-human acceptance criteria: a listed human must put the confirmed - [ ] criteria in the body or post them as their own comment, or, on an issue a listed human filed, intake's newest refinement comment must hold them unedited. Admission takes the issue once each of those is met, when intake triages it again (.github/actions/resolve-config/admission-rules.json).

  5. melodic-automation-lanes commented on Nov 21, 2026

    @melodic-automation-lanes
    Contributor

    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: When the stale-body reason is the only reason a deferred issue is refused, admit.mjs sends it back to intake at needs-triage, as it already does for a report-delivered blocker, so the hourly triage sweep re-triages it against the current body without a person.

    Acceptance criteria

    • In admit.mjs decide, a deferred issue (admission-deferred, no verdict comment) whose only refusal is deferralLapsed's body-edited reason (kind stale, hold false, line 1318) returns the retriage wait verdict (the existing path at admit.mjs:1721 and 2809) and not a refusal.
    • That wait leaves no refusal comment and no status: confirmed park; admit() sets the issue to needs-triage and posts a retriage note (retriageNote, admit.mjs:1774) that names the body edit.
    • A deferred issue refused for the body edit together with any other reason (not-ready, held, area: security, a merged Refs pull request, an open blocker) is still refused as before.
    • A deferral refused because area: security was added still holds for a person (kind stale, hold true) and is not sent to retriage.
    • The refusal text in .github/scripts/lib/admission-comment.mjs:12 no longer promises a re-triage for a body-edit refusal that nothing starts, or the case no longer reaches that text.
    • admit.test.mjs replaces the case at admit.test.mjs:381 (refuses a deferred issue whose body was edited after its deferral) with a test of the retriage outcome, and adds a test that a body edit plus another reason still refuses.
    • No edited trigger is added to intake-reenter.yml or intake-triage.yml, so an ordinary edit starts no run.
    • The intake-triage test suite (admit.test.mjs and apply-verdict.test.mjs) passes.

    External state: none

    Dependencies

    • The hourly triage sweep that dispatches needs-triage issues must keep running; the retriage path relies on it (admit.mjs:2810, #​17011).

    Related work

    Sources read

    • context.json: issue body picks the admit.mjs fix and rejects a new edited trigger; intake's triage comment agrees.
    • .github/scripts/intake-triage/admit.mjs:1317: deferralLapsed adds the body-edited reason as kind stale, hold false.
    • .github/scripts/intake-triage/admit.mjs:1634: the deferral check runs only with no verdict and the DEFERRED label, and pushes the reason into refuse.
    • .github/scripts/intake-triage/admit.mjs:1667: an existing wait path already sends an issue back to intake by retriage.
    • .github/scripts/intake-triage/admit.mjs:2809: the retriage wait sets needs-triage and posts a note; the hourly triage sweep dispatches intake again.
    • .github/scripts/lib/admission-comment.mjs:12: the refusal text promises that admission takes the issue when intake triages it again.
    • .github/workflows/intake-triage.yml:38: triggers are issues opened and reopened plus workflow_dispatch only.
    • .github/workflows/intake-reenter.yml:57: triggers are issue_comment created and issues unlabeled, field_added and field_removed.

    Automated refinement by the lanes App. Intake triages the issue again with it.

  6. melodic-automation-lanes commented on Nov 21, 2026

    @melodic-automation-lanes
    Contributor

    {"v":1,"actor_id":337595936,"title_sha256":"91dc233c06c785c2c2512812474d387d8d7213e5ecbc70e3c71c677d77d365d7","body_sha256":"61f8ea82f9a5db80ee1975ad594f165e8cee7ed932e06f968a52cb2c046c6e03"}

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    Fields

    Work status

    agent-ready

    Priority

    medium

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions