Zurück zu Fähigkeiten

review-pull-request

pjt222
Aktualisiert 2 days ago
8 Ansichten
17
2
17
Auf GitHub ansehen
Andereai

Über

Diese Claude Skill führt umfassende GitHub Pull-Request-Reviews unter Verwendung der GitHub CLI durch, analysiert Diffs, Commit-Verläufe und CI/CD-Status. Sie bietet Feedback mit Schweregraden (blockierend/Empfehlung/Kleinigkeit/Lob) und reicht Reviews direkt über `gh pr review` ein. Nutzen Sie sie, wenn Sie zur Überprüfung eines PRs zugewiesen sind, für Selbstreviews vor Einholung von Feedback oder für Qualitätsaudits nach dem Mergen.

Schnellinstallation

Claude Code

Empfohlen
Primär
npx skills add pjt222/agent-almanac -a claude-code
Plugin-BefehlAlternativ
/plugin add https://github.com/pjt222/agent-almanac
Git CloneAlternativ
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/review-pull-request

Kopieren Sie diesen Befehl und fügen Sie ihn in Claude Code ein, um diese Fähigkeit zu installieren

Dokumentation

Review Pull Request

Review a GitHub pull request end-to-end — from understanding the change through submitting structured feedback. Uses gh CLI for all GitHub interactions and produces severity-leveled review comments.

When to Use

  • A pull request is ready for review and assigned to you
  • Performing a second review after the author addresses feedback
  • Reviewing your own PR before requesting others' review (self-review)
  • Auditing a merged PR for post-merge quality assessment
  • When you want a structured review process rather than ad-hoc scanning

Inputs

  • Required: PR identifier (number, URL, or owner/repo#number)
  • Optional: Review focus (security, performance, correctness, style)
  • Optional: Codebase familiarity level (familiar, somewhat, unfamiliar)
  • Optional: Time budget for the review (quick scan, standard, thorough)

Procedure

Step 1: Understand the Context

Read the PR description and understand what the change is trying to accomplish.

  1. Fetch PR metadata:
    gh pr view <number> --json title,body,author,baseRefName,headRefName,labels,additions,deletions,changedFiles,reviewDecision
    
  2. Read the PR title and description:
    • What problem does this PR solve?
    • What approach did the author take?
    • Are there any specific areas the author wants reviewed?
  3. Check the PR size and assess time required:
PR Size Guide:
+--------+-----------+---------+-------------------------------------+
| Size   | Files     | Lines   | Review Approach                     |
+--------+-----------+---------+-------------------------------------+
| Small  | 1-5       | <100    | Read every line, quick review       |
| Medium | 5-15      | 100-500 | Focus on logic changes, skim config |
| Large  | 15-30     | 500-    | Review by commit, focus on critical  |
|        |           | 1000    | files, flag if should be split       |
| XL     | 30+       | 1000+   | Flag for splitting. Review only the  |
|        |           |         | most critical files.                 |
+--------+-----------+---------+-------------------------------------+
  1. Review the commit history:
    gh pr view <number> --json commits --jq '.commits[].messageHeadline'
    
    • Are commits logical and well-structured?
    • Does the history tell a story (each commit a coherent step)?
  2. Check CI/CD status:
    gh pr checks <number>
    
    • Are all checks passing?
    • If checks are failing, note which ones — this affects the review

Got: A clear understanding of what the PR does, why it exists, how big it is, and whether CI is green. This context shapes the review approach.

If fail: If the PR description is empty or unclear, note this as the first piece of feedback. A PR without context is a review antipattern. If gh commands fail, verify you're authenticated (gh auth status) and have access to the repository.

Step 2: Analyze the Diff

Read the actual code changes systematically.

  1. Fetch the full diff:
    gh pr diff <number>
    
  2. For small/medium PRs, read the entire diff sequentially
  3. For large PRs, review by commit:
    gh pr diff <number> --patch  # full patch format
    
  4. For each changed file, evaluate:
    • Correctness: Does the code do what the PR says it does?
    • Edge cases: Are boundary conditions handled?
    • Error handling: Are errors caught and handled appropriately?
    • Security: Any injection, auth, or data exposure risks?
    • Performance: Any obvious O(n^2) loops, missing indexes, or memory issues?
    • Naming: Are new variables/functions/classes named clearly?
    • Tests: Are new behaviors covered by tests?
  5. Take notes as you read, classifying each observation by severity

Got: A set of observations covering correctness, security, performance, and quality for every meaningful change in the diff. Each observation has a severity level.

If fail: If the diff is too large to review effectively, flag it: "This PR changes {N} files and {M} lines. I recommend splitting it into smaller PRs for more effective review." Still review the highest-risk files.

Step 3: Classify Feedback

Organize observations into severity levels.

  1. Classify each observation:
Feedback Severity Levels:
+-----------+------+----------------------------------------------------+
| Level     | Icon | Description                                        |
+-----------+------+----------------------------------------------------+
| Blocking  | [B]  | Must fix before merge. Bugs, security issues,      |
|           |      | data loss risks, broken functionality.             |
| Suggest   | [S]  | Should fix, but won't block merge. Better           |
|           |      | approaches, missing edge cases, style issues that   |
|           |      | affect maintainability.                            |
| Nit       | [N]  | Optional improvement. Style preferences, minor      |
|           |      | naming suggestions, formatting.                    |
| Praise    | [P]  | Good work worth calling out. Clever solutions,      |
|           |      | thorough testing, clean abstractions.              |
+-----------+------+----------------------------------------------------+
  1. For each Blocking item, explain:
    • What's wrong (the specific issue)
    • Why it matters (the impact)
    • How to fix it (a concrete suggestion)
  2. For each Suggest item, explain the alternative and why it's better
  3. Keep Nits brief — one sentence is enough
  4. Include at least one Praise if anything positive stands out

Got: A sorted list of feedback items with clear severity levels. Blocking items have fix suggestions. The ratio should generally be: few Blocking, some Suggest, minimal Nit, at least one Praise.

If fail: If everything seems blocking, the PR may need to be reworked rather than patched. Consider requesting changes at the PR level rather than line-by-line comments. If nothing seems wrong, say so — "LGTM" is valid feedback when the code is good.

Step 4: Write Review Comments

Compose the review with structured, actionable feedback.

  1. Write the review summary (top-level comment):
    • One sentence: what the PR does (confirm understanding)
    • Overall assessment: approve, request changes, or comment
    • Key items: list Blocking issues (if any) and top Suggest items
    • Praise: call out good work
  2. Write inline comments for specific code locations:
    # Post inline comments via gh API
    gh api repos/{owner}/{repo}/pulls/{number}/comments \
      -f body="[B] This SQL query is vulnerable to injection. Use parameterized queries instead.\n\n\`\`\`suggestion\ndb.query('SELECT * FROM users WHERE id = $1', [userId])\n\`\`\`" \
      -f commit_id="<sha>" \
      -f path="src/users.js" \
      -F line=42 \
      -f side="RIGHT"
    
  3. Format feedback consistently:
    • Start each comment with the severity tag: [B], [S], [N], or [P]
    • Use GitHub suggestion blocks for concrete fixes
    • Link to documentation for style/pattern suggestions
  4. Submit the review:
    # Approve
    gh pr review <number> --approve --body "Review summary here"
    
    # Request changes (when blocking issues exist)
    gh pr review <number> --request-changes --body "Review summary here"
    
    # Comment only (when unsure or providing FYI feedback)
    gh pr review <number> --comment --body "Review summary here"
    

Got: A submitted review with clear, actionable feedback. The author knows exactly what to fix (Blocking), what to consider (Suggest), and what went well (Praise).

If fail: If gh pr review fails, check permissions. You need write access to the repo or to be a requested reviewer. If inline comments fail, fall back to putting all feedback in the review body with file:line references.

Step 5: Follow Up

Track the review resolution.

  1. After the author responds or pushes updates:
    gh pr view <number> --json reviewDecision,reviews
    
  2. Re-review only the changes that address your feedback:
    gh pr diff <number>  # check new commits
    
  3. Verify Blocking items are resolved before approving
  4. Resolve comment threads as issues are addressed
  5. Approve when all Blocking items are fixed:
    gh pr review <number> --approve --body "All blocking issues resolved. LGTM."
    

Got: Blocking issues verified as fixed. Review conversation resolved. PR approved or further changes requested with specific remaining items.

If fail: If the author disagrees with feedback, discuss in the PR thread. Focus on impact (why it matters) rather than authority. If disagreement persists on non-blocking items, yield gracefully — the author owns the code.

Validation Checklist

  • PR context understood (purpose, size, CI status)
  • All changed files reviewed (or highest-risk files for XL PRs)
  • Feedback classified by severity (Blocking/Suggest/Nit/Praise)
  • Blocking items have specific fix suggestions
  • At least one Praise included for positive aspects
  • Review decision matches feedback (approve only if no Blocking items)
  • Inline comments reference specific lines with severity tags
  • CI/CD checks verified (green before approval)
  • Follow-up completed after author's revisions

Pitfalls

  • Rubber-stamping: Approving without actually reading the diff. Every approval is an assertion of quality
  • Nit avalanche: Drowning the author in style preferences. Save nits for mentoring situations; skip them in time-sensitive reviews
  • Missing the forest: Reviewing line-by-line without understanding the overall design. Read the PR description and commit history first
  • Blocking on style: Formatting and naming are almost never blocking. Reserve Blocking for bugs, security, and data integrity
  • No praise: Only pointing out problems is demoralizing. Good code deserves recognition
  • Review scope creep: Commenting on code that wasn't changed in the PR. If pre-existing issues bother you, file a separate issue

Related Skills

  • review-software-architecture — System-level architecture review (complementary to PR-level review)
  • security-audit-codebase — Deep security analysis for PRs with security-sensitive changes
  • create-pull-request — The other side of the process: creating PRs that are easy to review
  • commit-changes — Clean commit history makes PR review significantly easier

GitHub Repository

pjt222/agent-almanac
Pfad: i18n/caveman-lite/skills/review-pull-request
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

Verwandte Skills

llamaguard

Andere

LlamaGuard ist Metas 7-8B-Parameter-Modell zur Moderation von LLM-Eingaben und -Ausgaben in sechs Sicherheitskategorien wie Gewalt und Hassrede. Es bietet eine Genauigkeit von 94-95 % und kann mit vLLM, Hugging Face oder Amazon SageMaker eingesetzt werden. Nutzen Sie diese Skill, um Inhaltsfilterung und Sicherheitsguardrails einfach in Ihre KI-Anwendungen zu integrieren.

Skill ansehen

cost-optimization

Andere

Diese Claude Skill unterstützt Entwickler bei der Optimierung von Cloud-Kosten durch Ressourcen-Dimensionierung, Tagging-Strategien und Ausgabenanalysen. Sie bietet einen Rahmen zur Senkung von Cloud-Ausgaben und zur Implementierung von Kosten-Governance für AWS, Azure und GCP. Nutzen Sie sie, wenn Sie Infrastrukturkosten analysieren, Ressourcen richtig dimensionieren oder Budgetvorgaben einhalten müssen.

Skill ansehen

quantizing-models-bitsandbytes

Andere

Diese Fähigkeit quantisiert LLMs auf 8-Bit- oder 4-Bit-Präzision mittels bitsandbytes und erreicht dabei eine Speicherreduzierung von 50–75 % bei minimalem Genauigkeitsverlust. Sie ist ideal für den Betrieb größerer Modelle mit begrenztem GPU-Speicher oder zur Beschleunigung von Inferenzvorgängen und unterstützt Formate wie INT8, NF4 und FP4. Die Fähigkeit integriert sich in HuggingFace Transformers und ermöglicht QLoRA-Training sowie 8-Bit-Optimierer.

Skill ansehen

dispatching-parallel-agents

Andere

Diese Claude-Fähigkeit verteilt mehrere Agenten, um drei oder mehr unabhängige Probleme gleichzeitig zu untersuchen und zu beheben. Sie ist für Szenarien konzipiert, die unabhängige Fehler umfassen, die ohne gemeinsamen Zustand oder Abhängigkeiten gelöst werden können. Die Kernfähigkeit ist die parallele Problemlösung, bei der pro unabhängigem Problembereich ein Agent zugewiesen wird, um die Effizienz zu maximieren.

Skill ansehen