agentleFS
Sign inStart free
y-miyazakiy-miyazaki/config/.agents/skills/github-issue-autofix/SKILL.md

github-issue-autofix

Skill1 starsChanged 8 days agoApache-2.0View on GitHub

Implement a GitHub Issue fix on a branch and open a PR (Fixes #N). Use in conversation when the user gives an Issue number or URL and asks to implement or fix it. Also the L2 loop entry skill after autofix label, repository_dispatch, or workflow_dispatch. Interactive default is survey only; implement the fix when the request asks for changes to be made (for example fix, apply, or update) or automation sets agent_edit_allowed in Constraints. Detect skips when an open/draft PR already closes the Issue. Agent must not fire repository_dispatch.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When intent is unclear, survey and offer to apply — edits are harder to undo than a report.

    line 62
  2. Very specific:

    When agent_edit_allowed is true, resolve agent_edit_target: on the interactive path use fix (this skill).

    line 64
  3. Very specific:

    IF agent_edit_allowed is false → emit survey shape.

    line 70
  4. Very specific:

    Do not fire repository_dispatch or start other loops from the Agent.

    line 74

≈ “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

UTILITY SKILL — Issue→PR autofix on the branch caller (not entity observe).

Input

  • Automation: detect JSON from detect_autofix.sh with result.issue_number. Read agent_edit_allowed / agent_edit_target from ## Constraints.
  • Interactive: Issue number (or URL resolved to a number) plus optional failure context; there is no detect JSON. Read the Issue with read-only gh issue view <n> --json number,title,body,state,labels,comments. agent_edit_allowed follows the edit gate in Workflow: survey unless the request asks for changes to be made.

Output Specification

Report per common-output-format.md. On the automation path, follow category-automation-envelope.md. Apply shape: PR body includes the literal Fixes #<N>. Survey shape: reference the Issue as Refs #<N> and never use a closing keyword (Fixes, Closes, Resolves) — merging a survey PR must not close the Issue.

Execution Scope

USE FOR:

  • Implement a minimal fix for the Issue on the integration branch worktree when agent_edit_allowed: true
  • Synthesize PR body from assets/pr-body-template.md including Fixes #<N>

DO NOT USE FOR:

  • Issue triage labels/comments (github-issue-triage)
  • PR revision from review comments (github-pr-revise)
  • Calling repository_dispatch from the Agent

Reference Files Guide

  • common-checklist.md (always read)
  • common-output-format.md (always read)
  • category-automation-envelope.md (read on automation path)
  • category-pr-body-links.md (read when synthesizing PR body)
  • assets/pr-body-template.md (read when synthesizing apply PR body)
  • assets/pr-body-template-survey.md (read when synthesizing survey PR body)

Workflow

Resolve agent_edit_allowed before implementing:

Sourceagent_edit_allowed
Interactive — default: the request asks to look, check, triage, or survey, or only names the skillfalse — survey only; do not edit files
Interactive — the request asks for changes to be made (for example fix, apply, 修正して, 適用して, update, add)true — implement the fix for the Issue
Interactive — follow-up after a survey asking to applytrue — implement the surveyed fix candidates
Automation — ## Constraintsagent_edit_allowed: true or agent_edit_allowed: false from category-automation-envelope.md

When intent is unclear, survey and offer to apply — edits are harder to undo than a report.

When agent_edit_allowed is true, resolve agent_edit_target: on the interactive path use fix (this skill); on the automation path read agent_edit_target from ## Constraints. Do not branch on other caller metadata outside ## Constraints.

Delivery. Interactive: leave edits in the working tree and report them; commit, push, or open a PR only when the user explicitly asks for that step, because those change shared state that other people see and cannot be quietly taken back. Automation: do not push or open a PR yourself; the caller delivers the branch and composes the PR body from your report.

  1. Automation: read detect JSON; if skip: true or status: error, report the reason and stop. Interactive: resolve the Issue from the user's URL or number with gh issue view; if the Issue is closed, report that and stop. Check for an open or draft PR that already closes the Issue (for example gh pr list --state open --search "<n> in:body" and confirm a closing keyword for #<n> in its body); if one exists, report it as a skip and stop — same as the detect skip.
  2. On the automation path, read category-automation-envelope.md for Constraints and PR templates.
  3. IF agent_edit_allowed is false → emit survey shape; load assets/pr-body-template-survey.md at synthesis when on automation path; stop.
  4. ELSE IF agent_edit_allowed is true AND agent_edit_target is not fix → emit survey shape; note expected agent_edit_target: fix; stop.
  5. ELSE IF agent_edit_allowed is true AND agent_edit_target is fix → implement the minimal fix for Issue #N on the worktree.
  6. At synthesis, load the apply PR template; ensure body includes Fixes #<N> (interactive: the body is written for a PR the user asks you to open, or reported in chat).
  7. Do not fire repository_dispatch or start other loops from the Agent.

Error Handling

ConditionSeverityAction
Detect status: errorFatalStop; do not mutate repository files
Detect skip: trueInfoReport skip reason; do not open a PR
Interactive: Issue is closedInfoReport that the Issue is closed; do not edit files or open a PR
Interactive: an open or draft PR already closes the IssueInfoReport it as a skip with the PR number; do not open another PR
agent_edit_allowed: falseInfoSurvey only; do not edit files
agent_edit_target not fix when edits expectedRecoverableSurvey only; note mismatch in Overview
Validation tooling missing after editsRecoverableDefer in report; do not claim passing checks

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