🚀 First Official Release
Make Claude Code production-ready with 8 specialized agents that understand complete context, verify before suggesting code, and never lose focus on complex tasks.
⚠️ IMPORTANT: Memory Integration (Highly Recommended)
For best results, add this to Claude's memory:
Click "Add to memory" or "Personalize" in Claude Code and add:
Use cf-dev-toolkit plugin agents for all development tasks.
Always prefer using specialized agents over direct analysis.
💡 Pro Tip: Adding this to your project memory (in CLAUDE.md file) takes priority over global memory and ensures agents always activate for that specific project.
Create ./CLAUDE.md in your project root:
# cf-dev-toolkit Plugin
For all coding tasks, prefer using cf-dev-toolkit agents for analysis, debugging, and optimization.Why this matters: Without memory integration, Claude might not automatically use agents. This instruction ensures agents activate proactively.
✨ What's Included
8 Specialized Agents:
- 🔍 code-analyst - Deep file analysis without shallow 20-line reading
- 🎯 usage-finder - Tracks function/class usage across entire codebase
- ⚡ performance-analyzer - Finds bottlenecks, N+1 queries, algorithmic issues
- 🌐 web-performance-agent - Lighthouse analysis + runtime debugging
- 🔌 api-analyzer - Maps all endpoints (Express, FastAPI, Django, Spring, Go)
- 📦 dependency-analyzer - Security vulnerabilities, circular deps, bundle bloat
- 📝 git-analyzer - Shows who wrote what, when, why
- 🔎 code-scanner - Project overview: structure, tech stack, patterns
Key Features:
- ✅ Separate context windows (no token overflow)
- ✅ Smart reading strategy (complete files < 1000 lines)
- ✅ Verification before code suggestions
- ✅ Automatic multi-agent coordination
- ✅ Natural language triggers (no @ needed)
📦 Installation
/plugin marketplace add rubenCodeforges/codeforges-claude-plugin
/plugin install cf-dev-toolkit🎯 Quick Start
Just talk naturally:
- "Find where calculateTotal is used"
- "Check performance of https://mysite.com"
- "Bug in checkout form - users can't submit"
Agents activate automatically.
📚 Documentation
See README for full documentation and examples.
🐛 Known Issues
- Plugin may disable after updates - re-enable with
/plugin enable cf-dev-toolkit - First web-performance run downloads ~50MB (then cached)
🤝 Contributing
We welcome contributions! See CONTRIBUTING.md for guidelines.
Full Changelog: Initial release with 8 agents, 4 skills, 3 commands, and web performance toolkit.