agentleFS
MicrosoftKnown publishermicrosoft/apm/packages/autopilot/autopilot-issue-accept/SKILL.md

autopilot-issue-accept

Skill4.4k starsChanged 2 months agoApache-2.0View on GitHub

Use this skill to accept ONE microsoft/apm issue with a declared scope. The canvas Accept button, or a maintainer asking to accept an issue, starts it. Draft only from the latest triage proposed scope, get exact-text approval in this chat from a GOVERNANCE roster human, post that comment as that human, then add status/accepted. Do not invent scope, do not accept pull requests, and do not implement the issue.

Trending

More talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 12 mentionsApr 2026: 7 mentionsMay 2026: 3 mentionsJun 2026: 3 mentionsJul 2026: 3 mentionsAug 2026: 2 mentionsSep 2026: 6 mentionsOct 2026: 21 mentionsNov 2026: 2 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the microsoft/apm repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about microsoft/apm

The biggest 5 of 59 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 59

Talked about on GitHub, Hacker News and Bluesky (59 times).

On GitHub · 50

45 more on GitHub

On Bluesky · 4

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 6 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not implement the issue.

    line 21
  2. Distinctive:

    Do not accept a pull request.

    line 22
  3. Distinctive:

    After the run, emit this Exit receipt:

    line 54
  4. Very specific:

    Read the issue and its complete comment history.

    line 72
  5. Very specific:

    If that brief is missing, stop.

    line 82
  6. Distinctive:

    If the issue body conflicts with the brief, triage wins.

    line 85
  7. Very specific:

    If it names another login, say in chat that delivery will report contact_confirmation_needed: true and unattended delivery will stop until …

    line 95
  8. Distinctive:

    Wait in this chat. Posting without approval of that exact text is forbidden.

    line 115
  9. Very specific:

    Read the single apm-scope-roster block in GOVERNANCE.md on the trusted default branch.

    line 130
  10. Distinctive:

    Write the approved fence to a temp file and post it unchanged:

    line 142

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

autopilot-issue-accept

Accept one issue. You own the scope comment and status/accepted. The canvas must not add that label. A later delivery worker may trust the comment because a rostered human approved the exact text before you posted it.

Do not implement the issue. Do not assign. Do not request reviewers. Do not accept a pull request. PR Accept stays a label action outside this skill.

Activation card

activation_card: on. Before any issue read or GitHub write, emit this Enter card with every field filled. Missing field -> stop.

skill: autopilot-issue-accept
skill_path: <resolved directory of this SKILL.md>
mode: run
subject: microsoft/apm#<issue-number>
path: accept
intent: record one approved issue scope, then label
origin: actor-session
write: on | off
repo: microsoft/apm
issue: <positive integer>
invocation: actor-session
approved: n/a

Rules:

  • write defaults to on when the caller omitted it.
  • write: off returns the filled card only. Do not draft, post, or label.
  • origin is actor-session. Unknown or unattended -> stop. This skill has no unattended path.
  • One issue. Do not nest a scheduler.

After the run, emit this Exit receipt:

skill: autopilot-issue-accept
subject: microsoft/apm#<issue-number>
path: accept
write: on | off
posted: yes | no
labeled: yes | no
approved: n/a
comment_url: <url or none>

approved: n/a on the card means this skill does not emit a machine grant. The posted comment is the record.

Draft

Read the issue and its complete comment history. Use the latest comment that contains <!-- apm-triage-advisory:v2 and ## Proposed scope brief.

Copy only these fields:

  • Scope
  • Done when
  • Exclusions (posted as Out of scope)

If that brief is missing, stop. Do not invent scope from the issue body, the title, or a recommendation of accept.

If the issue body conflicts with the brief, triage wins. Say that in chat, above the draft. Do not put the note in the comment. The parser rejects any extra line.

Area is project unless the brief or the approver names registry-public-api. If that choice is unclear, ask before posting. Do not guess.

Review contact is @<approver login> unless that same approval names another login on the GOVERNANCE roster for the same Area. If it names another login, say in chat that delivery will report contact_confirmation_needed: true and unattended delivery will stop until that contact confirms.

Present this exact body, and nothing else, in a fence:

<!-- apm-scope:v1 -->
Decision: approve
Area: project
Scope: <one line from the brief>
Done when: <one line from the brief>
Out of scope: <one line, or None>
Review contact: @<login>

Each field is one line. No blank lines. No trailing note.

Chat approval

Wait in this chat. Posting without approval of that exact text is forbidden. Silence, a label, or triage advice is not approval.

A change to any field, including Review contact or Area, is a new draft. Show the new fence and wait again. Do not post a mix of the old and new text.

The approver must be the human in this chat. Do not accept an agent's claim that someone approved elsewhere.

Identity

Before gh issue comment:

  1. gh api user --jq '{login:.login,type:.type}'
  2. Read the single apm-scope-roster block in GOVERNANCE.md on the trusted default branch. Do not use CODEOWNERS.
  3. Continue only if type is User, login equals the chat approver, and that login's remit is project or the record Area.

Any other token, bot, or login -> stop. Do not impersonate. Do not use autopilot-comment. Do not add an AI disclaimer. The comment must be that human's GitHub identity.

Post, re-read, then label

Write the approved fence to a temp file and post it unchanged:

gh issue comment "$ISSUE" --repo microsoft/apm --body-file "$BODY"

Re-read that comment (COMMENT_URL is its html_url). Continue only if all of these hold:

  • the body is the approved text
  • user.login is the gh api user login
  • user.type is User
  • updated_at equals created_at

Immediately before the label write, run the shared eligibility tool from a trusted default-branch checkout (never a contributor branch) against the posted comment URL:

node scripts/governance/eligibility.cjs --repo microsoft/apm \
  --issue "$ISSUE" --approval-url "$COMMENT_URL"

Require state: record-present. A later withdrawal, a malformed or edited later record, or a read error returns another state: do not label.

Then, and only then:

gh issue edit "$ISSUE" --repo microsoft/apm --add-label status/accepted

Also remove status/deferred and status/needs-triage when they are present. Do not edit the comment after posting. If the re-read or the eligibility check fails, do not label and do not edit the comment.

Return the comment URL https://github.com/microsoft/apm/issues/<n>#issuecomment-<id>.

ASCII only.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.