The ultimate framework for supercharging your Claude Code workflow. Treat Claude as your brilliant, amnesiac coding partner and build amazing things together.
- Structured Memory Management - Never lose context again with the memory-bank system
- Checklist-Driven Development - Transform requirements into executable Claude prompts
- YOLO Mode - All permissions maxed out for ultimate productivity
- Large Codebase Support - Map and index massive projects with knowledge graphs
- Cross-AI Validation - Get second opinions from other AI models
- Session Management - Seamless handoffs between coding sessions
- Parallel Execution - Leverage multi-agent workflows for speed
This framework includes settings.json with ALL PERMISSIONS ENABLED. This means:
- Full filesystem access
- Unlimited token usage
- Parallel execution by default
- Auto-confirmations enabled
Use responsibly in development environments only!
# Install globally
npm install -g claude-code-framework
# Create a new project
claude init my-awesome-project --template web-app --yolo
# Start a session
cd my-awesome-project
claude session start
# Get your next task
claude task nextClaude Code is like a brilliant expert with short-term memory loss. This framework builds an external brain through:
- CLAUDE.md - Core rules and constraints
- PLAN.md - Checklist of executable prompts
- memory-bank/ - Persistent context storage
- Planning - Generate detailed implementation plans
- Implementation - Execute tasks one at a time
- Review - Cross-check with other AIs
- Session Management - Clean handoffs between sessions
claude init [name] # Initialize new project
claude doctor # Diagnose setup issues
claude config set <key> <value> # Configure settingsclaude plan generate <requirements.md> # Generate implementation plan
claude task next # Show next task
claude task complete # Mark current task done
claude session start --restore-context # Begin coding session
claude session end --update-memory # End session cleanlyclaude map-codebase --parallel-agents # Map large codebases
claude context compact # Optimize context usage
claude review --ai gemini --security # Cross-check codeyour-project/
├── CLAUDE.md # Core rules (don't delete!)
├── PLAN.md # Your implementation checklist
├── memory-bank/ # Claude's external brain
│ ├── projectBrief.md
│ ├── techContext.md
│ ├── systemPatterns.md
│ ├── activeContext.md
│ └── progress.md
├── .claude/ # Framework config
│ └── config.json
└── settings.json # YOLO mode permissions
- One Task at a Time - Focus prevents context pollution
- Commit Often - Lock in progress with Git
- Review Like PRs - Reject bad code entirely
- Update Memory - Always update context before ending sessions
- Cross-Check Plans - Use external AIs to validate
{
"yolo": {
"confirmations": {
"skipAll": true,
"autoYes": true
}
}
}The framework supports multiple MCP servers:
- basic-memory - Simple note-based memory
- graphiti - Advanced knowledge graphs
- sequential-thinking - Deep analysis mode
- puppeteer - Browser automation
Add these to prompts for enhanced capabilities:
ultrathink- Deep analysis modesub-task with agents- Parallel processing
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
MIT License - See LICENSE file for details.
Built on the shoulders of giants and the collective wisdom of the Claude Code community.
Remember: Claude Code is your brilliant partner who just happens to forget everything between sessions. This framework is their memory palace. Use it wisely! 🧠✨