code-review

Code Review (PR-Based, Elite/Rigorous)

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "code-review" with this command: npx skills add v1-io/v1tamins/v1-io-v1tamins-code-review

Code Review (PR-Based, Elite/Rigorous)

Perform a staff-level PR review using full repo context, then post the review to GitHub.

Usage

/code-review <PR_URL_or_NUMBER>

Examples:

/code-review https://github.com/your-org/your-repo/pull/123 /code-review 123

What It Does

  1. Gather PR Context

gh pr view <PR> --json title,body,author,baseRefName,headRefName,commits,files,labels,additions,deletions gh pr view <PR> --comments

Extracts:

  • What the PR claims to do (title/body/commits)

  • Where it touches (services, shared libs, frontend)

  • Risk flags: auth, migrations, background jobs, concurrency, data integrity

  1. Get the Diff

gh pr diff <PR> # Full diff gh pr diff <PR> --name-only # File list

  1. Build Repo-Aware Understanding

For each meaningful changed file:

  • Reads surrounding code (imports, callers, contracts)

  • Searches repo-wide for usage patterns: rg "<Symbol>" -n

  • Aligns with project architectural patterns

  1. Review Systematically

Correctness & Logic:

  • Boundary cases, idempotency, error paths

Concurrency & Async:

  • No blocking calls in async contexts

  • LangChain/LangGraph: prefer .ainvoke() over .invoke()

Security & Privacy:

  • No hardcoded secrets, input validation, no PII in logs

Performance & Cost:

  • No N+1 queries, no unbounded queries

Observability:

  • Logging conventions followed

Database/Migrations:

  • Models registered, updated_at updated, safe migrations

Tests:

  • Cover the "why", no flaky patterns
  1. Produce Outputs

Output A: Main Summary Comment

  • PR Summary (3-6 bullets)

  • Key Findings (severity-ranked: Critical/Medium/Low/Nit)

  • Confidence Score (X/5 with reasoning)

  • File-level confidence table

  • Mermaid sequence diagram (if non-trivial)

Posted via:

gh pr review <PR> --comment -b "<BODY>"

Or for must-fix issues:

gh pr review <PR> --request-changes -b "<BODY>"

Output B: Per-File Comments One comment per file with actionable suggestions:

Review: path/to/file (Importance: Medium)

  • Issue: <short title>
  • Why it matters: <tie to project patterns>
  • Suggested change: <what to do>
  • Test: <test to add if applicable>

Posted via:

gh issue comment <PR_URL> --body "<BODY>"

Confidence Scoring Guide

  • 5/5: Trivial change, well-tested, no risk

  • 4/5: Standard change, good coverage, minor concerns

  • 3/5: Non-trivial change, needs attention in specific areas

  • 2/5: Significant concerns, missing tests, risky patterns

  • 1/5: Likely bugs, security issues, or major problems

Reduce score for: migrations, auth/permissions, concurrency, broad refactors, missing tests.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

interview-me

No summary provided by upstream source.

Repository SourceNeeds Review
General

write-tests

No summary provided by upstream source.

Repository SourceNeeds Review
General

complexity

No summary provided by upstream source.

Repository SourceNeeds Review