Skip to content

osocode/code-review-packs

Repository files navigation

Code Review Packs

Versioned, AI-tool-agnostic code review frameworks for development teams.

What is a Code Review Pack?

A Code Review Pack is a self-contained set of review guidelines, checklists, and AI tool configurations for a specific technology stack. Packs enable consistent, thorough code reviews whether performed by humans, AI assistants in IDEs, or automated CI pipelines.

Available Packs

Pack Description Status
python-azure-ai-agent Python AI agents on Azure AI Foundry with Microsoft Agent Framework v0.0.1

Quick Start

Option 1: CLI Installation

pip install code-review-pack

# Initialize a pack in your project
code-review-pack init --pack python-azure-ai-agent

# Or run interactively
code-review-pack init

# Run a local code review on your changes
code-review-pack review

# Review only staged changes
code-review-pack review --staged

Option 2: Manual Installation

  1. Copy the desired pack from packs/<pack-name>/ to your repository
  2. Copy AI tool configurations to appropriate locations:
    • Windsurf: .windsurf/rules/ and .windsurf/workflows/
    • Cursor: .cursorrules and AGENTS.md to repo root
    • Claude Code: CLAUDE.md to repo root
  3. Copy GitHub Action to .github/workflows/ for CI reviews

Supported AI Tools

  • Windsurf: Rules and Cascade Workflows
  • Cursor: Rules and Agent definitions
  • Claude Code: CLAUDE.md project instructions
  • Claude Desktop/Web: System prompts and checklists
  • GitHub Actions: Automated PR reviews via Claude API

Review Dimensions

Each pack reviews code across these dimensions:

  1. Correctness - Logic, edge cases, error handling
  2. Readability - Clarity, naming, documentation
  3. Architecture - Structure, boundaries, patterns
  4. Language Patterns - Idiomatic usage, anti-patterns
  5. Security - Traditional AppSec concerns
  6. AI Security - Prompt injection, tool safety, data leakage
  7. Dependencies - Necessity, versions, supply chain
  8. Performance - Efficiency, scalability, resource usage
  9. Operations - Observability, deployment, configuration
  10. Tests - Coverage, quality, maintainability

Documentation

Contributing

Contributions welcome! See CONTRIBUTING.md for guidelines.

License

Apache License 2.0 - see LICENSE

About

Code Review Packages to setup local and GitHub Action code reviews with common AI enabled IDEs

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages