Git PR Reviewer - Setup Guide¶
Source: onewave-ai/claude-skills (239+ installs) Category: Engineering / Code Review Quality Tier: 🟡 Beta
Automated PR review skill that provides a structured checklist covering security, code quality, performance, and testing. Designed to be invoked when reviewing pull requests, checking code changes, or analyzing diffs before merge.
Installation¶
npx skills add onewave-ai/claude-skills --skill git-pr-reviewer
Prerequisites¶
| Requirement | Details |
|---|---|
| Git | Standard git CLI with access to the target repository |
| Shell access | Read, Grep, Glob, Bash tools available |
Key Capabilities¶
Security Review¶
Scans for hardcoded credentials, missing input validation, SQL injection patterns, XSS vulnerabilities, and authentication/authorization gaps.
Code Quality Analysis¶
Checks function size, code duplication, naming conventions, error handling patterns, and dead code.
Performance Audit¶
Identifies unnecessary database queries, inefficient algorithms, missing caching, and memory leak patterns.
Test Coverage Assessment¶
Verifies new code has tests, edge cases are covered, and tests are meaningful rather than coverage-only.
Quick Start¶
# Review current branch against main
git diff main...HEAD --stat
git diff main...HEAD
The skill analyzes all modified, added, and deleted files against its review checklist, producing a structured markdown summary with Critical / Warnings / Suggestions tiers and an Approval Status.
Verification¶
# Check the skill is installed
npx skills list | grep git-pr-reviewer
# Run on current branch
git diff main...HEAD | head -50
Notes¶
- Produces structured output: Overview → Issues Found (Critical/Warnings/Suggestions) → Approval Status
- Uses
git diff main...HEADby default; configurable base branch - Complements
sickn33/antigravity-awesome-skills@git-pr-review(PR description generator) anddavila7/claude-code-templates@create-pr(PR creation) - Quality tier 🟡 Beta: 239 installs, active development
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.