agentleFS
Sign inStart free
MicrosoftKnown publishermicrosoft/apm/packages/autopilot/autopilot-comment/SKILL.md

autopilot-comment

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

Use this skill to post or patch ONE GitHub comment for a microsoft/apm autopilot run. It owns public vs debug body assembly and the AI disclaimer footer. Never labels, assigns, requests reviewers, or merges. Autopilot skills that write comments must load this skill and must not call gh comment APIs themselves.

1 other project uses a version of this.

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.

Most recently taken up by

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After the write (or dry-run), emit this Exit receipt:

    line 65
  2. Distinctive:

    If the footer is already the last lines, do not duplicate it.

    line 92
  3. Distinctive:

    Do not assign, request reviewers, merge, or edit labels.

    line 109
  4. Distinctive:

    Do not compose schedulers or other workers.

    line 111

≈ “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-comment -- Sole GitHub comment writer

Any autopilot skill that would write a GitHub comment loads this skill in the same session and stops if the card is missing. Do not spawn a nested session. Do not comment from the caller.

Activation card

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

skill: autopilot-comment
skill_path: <resolved directory of this SKILL.md>
mode: run
subject: microsoft/apm#<number>
path: comment
intent: post one GitHub comment | patch one GitHub comment
origin: unattended | actor-session
write: on | off
debug: off | on
repo: microsoft/apm
target_kind: issue | pull
target: <positive integer>
source_skill: <canonical autopilot skill name>
comment_id: none | <positive integer>
invocation: agentic-workflow | actor-session
receipt: <HTML comment watermark line>
public_body: <author-facing prose, no JSON, no footer>
debug_body: <full internal template without JSON fence, or none>

Rules:

  • write defaults to on when the caller omitted it.
  • write: off returns the assembled body only. Do not post or patch.
  • write: on posts or patches exactly one comment. Never labels, assigns, requests reviewers, or merges.
  • debug defaults to off when omitted or unknown. Omitted debug is not a missing-field stop.
  • debug: off -> HTML receipt plus public_body. Do not post internal headings from debug_body.
  • debug: on -> first line [i] Skill debug is on. then receipt plus debug_body (or public_body if debug_body is none).
  • comment_id: none with intent post. A numeric comment_id with intent patch. Mismatch -> stop.
  • source_skill is the caller (for example autopilot-issue-triage-worker). Never autopilot-comment.
  • origin fail-closed unknown -> unattended.
  • Strip any JSON fence from bodies before assembly. Never post JSON on GitHub.

After the write (or dry-run), emit this Exit receipt:

skill: autopilot-comment
subject: microsoft/apm#<number>
path: comment
write: on | off
debug: off | on
posted: yes | no
patched: yes | no
comment_id: <id or none>
source_skill: <canonical autopilot skill name>
approved: n/a

Assemble

  1. Drop JSON fences and triage-recommendation blocks.
  2. Build the body from debug as above. ASCII only.
  3. Append this exact footer, substituting source_skill:
---

Generated by <source_skill>. This comment is AI-generated and may contain errors.
  1. If the footer is already the last lines, do not duplicate it.

Write

Actor-session / Copilot App / Cloud / Remote:

  • issue + post: gh issue comment <target> --body-file ...
  • pull + post: gh pr comment <target> --body-file ...
  • patch: gh api -X PATCH /repos/microsoft/apm/issues/comments/<comment_id>

Agentic Workflow: safe-outputs.add-comment for post. Patch is not available; stop if intent is patch.

One comment per activation. No progress comments.

Hard nos

  • Do not assign, request reviewers, merge, or edit labels.
  • Do not implement.
  • Do not compose schedulers or other workers.
  • 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.

Check your setup, free