Build & Code / D03
FreeReview a pull request
Find behavioral risks, missing tests, unclear changes, and high-signal review comments.
For
Teams
Time per use
10 min
Format
.md and .skill
How to use it
- 1.Open ChatGPT or Claude.Either works. The skill is just text.
- 2.Copy this skill from the free shelf.One click; no install, no setup.
- 3.Paste it as your first message.The assistant now knows how to do this one job.
- 4.Give it your specifics, get the result.Roughly 10 min, every time you need it.
Skill filed03-review-a-pull-request.skill.md1.6 KB
Run once
Advanced
Install permanently ↓Mobile apps: if ChatGPT or Claude opens blank, tap the message box and paste. Skillbook copies first.
Mobile apps: if ChatGPT or Claude opens blank, tap the message box and paste. Skillbook copies first.
Fill the blanks first.
These fields update the skill preview and the ChatGPT/Claude buttons instantly.
IncludePaste the diff, PR summary, changed files, or relevant snippets.
IncludeWhat should this change do?
IncludeTests run, screenshots, QA notes, or missing test context.
IncludeAuth, payments, data migration, mobile layout, API contracts, accessibility, performance, etc.
IncludeFramework, conventions, relevant files, users affected, rollout plan, or constraints.
Install as agent behavior
Install this as reusable agent behavior.
These versions preserve the blanks so the assistant asks for details every time, instead of hard-coding today's trip or task.
# Review a pull request
You are reviewing a pull request like a careful senior engineer. Prioritize bugs, behavioral regressions, missing tests, accessibility issues, security risks, and confusing contracts. Do not nitpick style unless it hides a real problem.
## What I will give you
- PR description, diff, or file summaries.
- Intended behavior.
- Test evidence if available.
- Risk areas I already suspect.
## What I want back
**1. Findings first.** Ordered by severity. For each finding, include:
- Severity: P0, P1, P2, or P3.
- File or area.
- The issue.
- Why it matters.
- A concrete fix or question.
**2. Missing tests.** Specific tests that would catch the riskiest behavior.
**3. Open questions.** Things I would ask before approving.
**4. Approval summary.** One short paragraph: approve, approve with nits, or request changes.
## Rules
- Lead with risks, not compliments.
- Do not assume the code works because it compiles.
- If the diff is too large, identify the riskiest paths first.
- Separate confirmed problems from suspicions.
- Do not ask for broad rewrites unless the current shape creates real risk.
PR or diff: {{diff_or_summary||Paste the diff, PR summary, changed files, or relevant snippets.}}
Intended behavior: {{intended_behavior||What should this change do?}}
Test evidence: {{test_evidence||Tests run, screenshots, QA notes, or missing test context.}}
Risk areas: {{risk_areas||Auth, payments, data migration, mobile layout, API contracts, accessibility, performance, etc.}}
Project context: {{project_context||Framework, conventions, relevant files, users affected, rollout plan, or constraints.}}On the house. Run it in ChatGPT or Claude, or install it as agent behavior.
↓ Download .mdOther things on the Build & Code shelf