code-review
Code Review
Run automated code review using cubic review with amp review as fallback.
When to Use
- Review code changes before committing
- Check PR quality
- Analyze code for issues
- Any request to "review code", "run code review", or "check quality"
Workflow
1. Check What to Review
Review staged changes:
git diff --staged --stat
Review specific files:
git diff path/to/file
Review current branch vs main:
git diff main...HEAD --stat
2. Run Code Review
Primary: cubic review
cubic review
With options:
cubic review --staged
cubic review path/to/file.ts
cubic review --diff main...HEAD
3. Fallback if cubic fails
If cubic review fails or is unavailable, use amp review:
amp review
With options:
amp review --staged
amp review path/to/file.ts
Complete Flow with Fallback
# Try cubic first
cubic review
# If cubic fails, fall back to amp
amp review
What They Check
cubic review typically analyzes:
- Code quality and best practices
- Potential bugs and issues
- Performance concerns
- Security vulnerabilities
- Style violations
amp review typically provides:
- Similar code quality checks
- Alternative analysis engine
- Different rule sets and heuristics
Review Scenarios
Review all staged changes:
cubic review --staged || amp review --staged
Review before committing:
git add .
cubic review --staged || amp review --staged
# Fix issues, then commit
git commit -m "fix: resolve review issues"
Review specific file:
cubic review src/auth.ts || amp review src/auth.ts
Review branch changes:
cubic review --diff main...HEAD || amp review --diff main...HEAD
Interpreting Results
- Address critical issues first (security, bugs)
- Consider warnings for maintainability
- Style issues can be auto-fixed in many cases
- Not all suggestions need to be implemented
Best Practices
- Run review before committing
- Fix critical issues immediately
- Use reviews to learn patterns
- Don't blindly accept all suggestions
- Consider the context of changes
- Run in CI/CD for automated checks
More from pc-style/pc-skills
blog-post
Create pcstyle.dev developer blog posts in the dual-author (ME MYSELF + MY AI AGENT) format, including MDX structure, custom components, and API/CLI publishing steps.
4load-handoff
Load a handoff file and proceed with the specified tasks. Use when starting a new session with a HANDOFF.md file present, or when the user asks to continue from a handoff. Reads the handoff context and immediately begins executing the next steps.
3create-handoff
Create a handoff message to continue work in a new AI session. Use when the user wants to save context for later, create a handoff document, or prepare work for continuation in a fresh session. Captures relevant files, context, and goals so the next session can start immediately without rediscovering the codebase.
3git-commit-push
Stage changes, commit with conventional commits, and push to origin. Use when the user wants to commit and push in one go, save changes to GitHub, or after completing work on a branch. Combines staging, committing, and pushing into a single workflow.
3generate-agents-md
Generate AGENTS.md file for a codebase to help AI agents understand the project structure, build commands, testing, and conventions. Use when the user asks to create an AGENTS.md, initialize agent documentation, or analyze a codebase for agent context. Also use when setting up a new repository for AI-assisted development.
3linear-status-check
Audit Linear project status against GitHub repository state to identify sync issues and generate actionable next steps. Use for checking if Linear issues match actual code implementation, finding discrepancies between marked status and reality, and creating detailed prompts for coding tools like Claude Code.
3