Index of every page on this site: llms.txt
All skills
vinta avatar

/review-prs

@4c3cb15
by Vinta Chenvinta/awesome-python336k stars
29,136
Run it yourself

Triage open PRs — screen from the diff, delegate the admission judgment to audit-the-list, then merge or close on GitHub. Fire when the maintainer asks to review PRs, process the PR queue, or judge whether a specific PR should be merged.

Use this Skill: https://skilld.dev/gh/vinta/awesome-python/review-prs

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈63 tokens for metadata: the name and description. ≈1.1k when used: this file.

Description uses 3.1% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈250 characters, or 3.1%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Review PRs

A PR review turns the open queue into terminal states: merged, closed, or explicitly parked. The rules live in CONTRIBUTING.md (Quality Requirements, Admission, Review Process, Automatic Rejection) — this skill is the workflow that applies them, not a second copy. Every PR named in chat is a PR link, [#<number>](https://github.com/vinta/awesome-python/pull/<number>), so the maintainer can open it from the terminal.

1. Fetch

Arguments name specific PRs; otherwise take the queue: gh pr list --repo vinta/awesome-python --limit 10 --json number,title,author,url,body,files,mergeable,mergeStateStatus. Fetch all diffs in parallel: gh pr diff <number> --repo vinta/awesome-python.

Sort the batch: entry additions continue; anything else (typo fixes, website changes, docs) is out of scope — report it under "needs human" and touch nothing, so it resurfaces every run until a human acts. Done when every PR is sorted and has its diff.

2. Screen

Apply the Automatic Rejection rules that the diff and PR metadata answer without judgment. The recently-closed-duplicate rule needs a lookup: gh pr list --repo vinta/awesome-python --state closed --search "<project name>" --limit 10; a project under docs/audit-logs.md "Rejected Additions" also counts as previously rejected unless the PR brings evidence newer than its line. A screened-out PR goes straight to step 4 as a close, with the rule it broke as its reason.

Resolve each survivor's target use case against the current README — a diff's context lines show the base the PR was written on, which may have changed since. A merge conflict rides to the Merge arm, which absorbs it. Done when every surviving PR names its target use case.

3. Judge

Group survivors by target use case — PRs proposing entries for the same use case compete for the same slots, so they ride one invocation. Per group, invoke the audit-the-list skill with arguments in this shape: "Judge proposed entries <names, each with its PR number> for the <section — subcategory> use case. Evidence and Verdicts steps only; report the verdicts back. No preview page, no README changes, no commits." The verdict for each PR is merge or close, grounded in the fetched evidence; when the use case is at cap, a merge verdict names the entry that leaves. An entry no current use case fits is a third outcome — a structure question: carry it to Act with evidence attached, and the maintainer decides there (mint the subcategory and merge, close, or leave open). Done when every survivor holds a verdict with a reason.

4. Act

  • Close: draft each closing comment — it states the reason and links CONTRIBUTING.md. Per batch of up to 4 closing PRs: print each draft in chat, its PR link as a label above a fenced block holding only the comment bytes, then AskUserQuestion over that batch — arms: close with the draft as printed, close without a comment, keep open. The chat print is the review copy; an option description only points at it. Keep a checklist of which verdicts have been asked; answers often arrive as custom text, and that text is the decision — an edited comment replaces the draft. Then gh pr close <number> --repo vinta/awesome-python --comment "<comment>", or a plain close.
  • Merge: a clean PR merges with gh pr merge <number> --repo vinta/awesome-python --merge. A conflicted one merges locally: git fetch origin pull/<number>/head, git merge FETCH_HEAD with the standard Merge pull request #<number> from <owner>/<headRef> message, resolving the conflict by placing the entry correctly — GitHub still marks the PR merged and the contributor keeps credit. Either way, reconcile the section per CONTRIBUTING before pushing: remove the entry the verdict displaced, fix the new entry's display name and Entry Ordering position, make test, commit. An add-only diff that displaces is the normal case — the removal is this step's job, not the contributor's. Done when every verdict has been adjudicated by the maintainer and its action executed, with every posted comment byte-identical to a printed draft or to the custom text the maintainer supplied. A kept-open PR resurfaces next run — that is its point.

5. Report

Summary table: PR link, verdict, action taken, plus the needs-human list. Done when every fetched PR ends in exactly one state — merged with its section reconciled, closed, or left open (kept open, needs-human, or structure question pending).

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 warning28d3 checks · Risk SAFE
  • Gen Agent Trust Hub28d

    The skill automates GitHub PR management for the vinta/awesome-python repository. It presents a low risk of indirect prompt injection because it processes untrusted pull request content from external contributors, though this is mitigated by mandatory human-in-the-loop approval steps before any destructive actions.

  • Socket1mo

    No alerts

  • Snyk1mo

    Risk: MEDIUM · 1 issue

Signed by skilld at 4c3cb15. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 days ago.

Activeupdated last month
argument-hint
[
  "PR numbers"
]

README badge

README badge for vinta/awesome-python/review-prs