CommunitySchreiben & Editierengithub.com

ChristianSavinoDP/Claude-Skills

Helper to get Claude to work better

Was ist Claude-Skills?

Claude-Skills is a Claude Code agent skill that helper to get Claude to work better.

Funktioniert mitClaude Code~Codex CLI~Cursor
npx skills add ChristianSavinoDP/Claude-Skills

Installed? Explore more Schreiben & Editieren skills: steipete/notion, affaan-m/seo, affaan-m/brand-voice · View all 6 →

In Ihrer bevorzugten KI fragen

Öffnet einen neuen Chat, in dem dieser Agent-Skill bereits geladen ist.

Dokumentation

PR Review

Procedure for reviewing a PR. The review rules (tone, comment categories, decision flow, what not to do) live in the Playbook under "PR Reviews"; this skill is the how. Do not restate the rules here, apply them. The Playbook's "Shared Standards" apply throughout (concise, no AI slop, verify do not assert, never fabricate).

Before reviewing

  1. Ask for the ticket first. Per the Playbook's mandatory first step, do not fetch the diff or analyze anything until you have the ticket. The ticket sets the work type (feature vs follow-up vs investigation) and acceptance criteria, which change how you review.
  2. Identify the PR: ask for the PR number/URL if not given.

Gather context

  • Fetch the diff and metadata with the GitHub CLI: gh pr view <pr> and gh pr diff <pr>.
  • For an investigation PR, also check docs/investigations/ for the expected format.
  • Read the changed files in full, not just the diff hunks, when behavior depends on surrounding code.

Review

Walk the diff file by file. Apply the Playbook's "PR Reviews" section:

  • Run the Playbook checklist (acceptance criteria, compile/test failures, callers updated, test coverage, evidence for investigations).
  • Classify each finding as blocking, nit, or question, per the Playbook.
  • Use the Playbook's decision flow to decide whether each observation is worth a comment.
  • Scope rules differ by ticket type (follow-up: resolve everything here; feature: refactors valid). Confirm the type from the ticket.
  • Per the Playbook's "Verify, do not assert": check claims yourself before repeating them. Run gh pr checks <pr> for CI, grep the repo for removed symbols. The PR description is the author's claim, not evidence.
  • Per the Playbook's "Do not silently drop existing behavior": an unconfirmed behavioral regression is a blocking finding, even if the author flagged it in the description.

Output

Follow the Playbook's "PR Reviews > Format". Output the review directly in the chat (not a file, unless the user asks for one), as a one-line verdict then findings. It is NOT a prose essay, a conversation log, or sections like "Context Gathered"/"Initial Review". Each finding is exactly: a location+code reference line, the comment in a copy-pasteable fenced block, then the why. Concretely, a finding looks like this:

### Blocking

`internal/adapters/users/adapter.go:281`
```go
_, err := u.providersClient.Update(ctx, request)
```

Comment (paste into the PR):

This drops the eventual-consistency retry these writes rely on. The Core->Users lag surfaces as NotFound, which the new transient-only policy will not retry, so the write fails immediately.


Why: AC #3 asks to confirm no path relies on retrying app-level errors; this path did, and the diff removes it without confirming the window is gone.
````

Repeat one such block per finding, grouped under `### Blocking` / `### Nits` / `### Questions` (omit empty groups). The fenced block after "Comment" is the only thing the user copies to GitHub; everything else is for navigation. If everything is good, just the verdict line.

## Posting (only if asked)

If the user wants comments posted rather than drafted in chat, confirm first, then use `gh pr review` / `gh pr comment`. Default to drafting in chat.

Verwandte Skills

steipete/notion

Notion CLI/API for pages, Markdown content, data sources, files, comments, search, Workers, and raw API calls.

community

affaan-m/seo

Audit, plan, and implement SEO improvements across technical SEO, on-page optimization, structured data, Core Web Vitals, and content strategy. Use when the user wants better search visibility, SEO remediation, schema markup, sitemap/robots work, or keyword mapping.

community

affaan-m/brand-voice

Build a source-derived writing style profile from real posts, essays, launch notes, docs, or site copy, then reuse that profile across content, outreach, and social workflows. Use when the user wants voice consistency without generic AI writing tropes.

community

affaan-m/crosspost

Multi-platform content distribution across X, LinkedIn, Threads, and Bluesky. Adapts content per platform using content-engine patterns. Never posts identical content cross-platform. Use when the user wants to distribute content across social platforms.

community

affaan-m/x-api

X/Twitter API integration for posting tweets, threads, reading timelines, search, and analytics. Covers OAuth auth patterns, rate limits, and platform-native content posting. Use when the user wants to interact with X programmatically.

community

affaan-m/content-engine

Create platform-native content systems for X, LinkedIn, TikTok, YouTube, newsletters, and repurposed multi-platform campaigns. Use when the user wants social posts, threads, scripts, content calendars, or one source asset adapted cleanly across platforms.

community