Claude-Flow is the ultimate multi-terminal orchestration platform that revolutionizes how you work with Claude Code. Coordinate multiple AI agents simultaneously, each working on different aspects of your project while sharing knowledge through an intelligent memory bank.
🔥 One command to rule them all:
npx claude-flow@latest- Deploy a full AI agent coordination system in seconds!
- ✅ Hundreds Concurrent Agents: Deploy up to 100+ AI agents simultaneously via BatchTool
- ✅ TypeScript Infrastructure Improvements: 91% reduction in compilation errors (379→32)
- ✅ Enhanced Test Framework: Parallel testing with comprehensive coverage
- ✅ Improved Build Process: Streamlined Deno compilation with dependency fixes
- ✅ Advanced Swarm Coordination: Multi-agent task distribution and monitoring
- Fixed import path issues and dependency management
- Enhanced error handling and type safety
- Improved test utilities and async operations
- Optimized parallel execution efficiency (71% faster)
- Maintained full backward compatibility
# Install and initialize with SPARC development environment
npx claude-flow@latest init --sparc
# Start the orchestration system
./claude-flow start --ui
# Deploy multiple agents with BatchTool
./claude-flow swarm "Build a REST API" --max-agents 5 --parallel# List all available SPARC modes
./claude-flow sparc modes
# Run specific development workflows
./claude-flow sparc run architect "design microservice architecture"
./claude-flow sparc run tdd "create comprehensive test suite"
./claude-flow sparc run security-review "audit authentication system"- Parallel Execution: Run up to 10 agents concurrently with BatchTool
- Smart Coordination: Intelligent task distribution and load balancing
- Memory Sharing: Persistent knowledge bank across all agents
- Real-time Monitoring: Live dashboard for agent status and progress
- 17 Specialized Modes: Architect, Coder, TDD, Security, DevOps, and more
- Workflow Orchestration: Complete development lifecycle automation
- Interactive & Non-interactive: Flexible execution modes
- Boomerang Pattern: Iterative development with continuous refinement
- System Health Dashboard: Real-time metrics and performance tracking
- Task Coordination: Dependency management and conflict resolution
- Terminal Pool Management: Efficient resource utilization
- Coverage Reports: Comprehensive test and code coverage analysis
# Initialize with full SPARC environment
npx claude-flow@latest init --sparc
# This creates:
# ✓ CLAUDE.md (SPARC-enhanced configuration)
# ✓ .roomodes (17 pre-configured SPARC modes)
# ✓ Local ./claude-flow executable
# ✓ Memory and coordination directories
# ✓ Claude Code slash commands# Install globally
npm install -g claude-flow@latest
# Initialize anywhere
claude-flow init --sparc# Clone and build from source
git clone https://github.com/ruvnet/claude-code-flow.git
cd claude-code-flow
deno task build# Check system status
./claude-flow status
# Start orchestration with UI
./claude-flow start --ui
# Spawn individual agents
./claude-flow agent spawn researcher --name "DataBot"
./claude-flow agent spawn coder --name "DevBot"# Deploy swarm for full-stack development
./claude-flow swarm "Build e-commerce platform" \
--strategy development \
--max-agents 5 \
--parallel \
--monitor
# BatchTool parallel development
batchtool run --parallel \
"./claude-flow sparc run architect 'design user auth'" \
"./claude-flow sparc run code 'implement login API'" \
"./claude-flow sparc run tdd 'create auth tests'" \
"./claude-flow sparc run security-review 'audit auth flow'"# Complete development workflow
./claude-flow sparc run ask "research best practices for microservices"
./claude-flow sparc run architect "design scalable architecture"
./claude-flow sparc run code "implement user service"
./claude-flow sparc run tdd "create comprehensive test suite"
./claude-flow sparc run integration "integrate all services"
./claude-flow sparc run devops "setup CI/CD pipeline"# Store and query project knowledge
./claude-flow memory store requirements "User auth with JWT"
./claude-flow memory store architecture "Microservice design patterns"
./claude-flow memory query auth
# Task coordination
./claude-flow task create research "Market analysis for AI tools"
./claude-flow task workflow examples/development-pipeline.json| Command | Description | Example |
|---|---|---|
init |
Initialize project with Claude integration | ./claude-flow init --sparc |
start |
Start orchestration system | ./claude-flow start --ui |
status |
Show system health and metrics | ./claude-flow status |
agent |
Manage AI agents and hierarchies | ./claude-flow agent spawn researcher |
swarm |
Advanced multi-agent coordination | ./claude-flow swarm "Build API" --parallel |
| Mode | Purpose | Example |
|---|---|---|
architect |
System design and architecture | ./claude-flow sparc run architect "design API" |
code |
Code development and implementation | ./claude-flow sparc run code "user authentication" |
tdd |
Test-driven development | ./claude-flow sparc run tdd "payment system" |
security-review |
Security auditing and analysis | ./claude-flow sparc run security-review "auth flow" |
integration |
System integration and testing | ./claude-flow sparc run integration "microservices" |
devops |
Deployment and CI/CD | ./claude-flow sparc run devops "k8s deployment" |
| Command | Description | Example |
|---|---|---|
memory store |
Store information in knowledge bank | ./claude-flow memory store key "value" |
memory query |
Search stored information | ./claude-flow memory query "authentication" |
task create |
Create and manage tasks | ./claude-flow task create research "AI trends" |
monitor |
Real-time system monitoring | ./claude-flow monitor --dashboard |
┌─────────────────────────────────────────────────────────┐
│ BatchTool Orchestrator │
├─────────────────────────────────────────────────────────┤
│ Agent 1 Agent 2 Agent 3 Agent 4 Agent 5 │
│ Architect │ Coder │ TDD │ Security │ DevOps │
├─────────────────────────────────────────────────────────┤
│ Shared Memory Bank & Coordination │
├─────────────────────────────────────────────────────────┤
│ Terminal Pool & Resource Management │
├─────────────────────────────────────────────────────────┤
│ Claude Code Integration Layer │
└─────────────────────────────────────────────────────────┘
- 🎛️ Orchestrator: Central coordination and task distribution
- 🤖 Agent Pool: Specialized AI agents for different domains
- 🧠 Memory Bank: Persistent knowledge sharing across agents
- 📊 Monitor: Real-time metrics and health monitoring
- 🔗 MCP Server: Model Context Protocol for tool integration
# Run full test suite with parallel agents
deno task test
# Run specific test categories
deno task test:unit # Unit tests
deno task test:integration # Integration tests
deno task test:e2e # End-to-end tests
# Generate coverage reports
deno task test:coverage- ✅ TypeScript Errors: Reduced from 379 to 32 (91% improvement)
- ✅ Build Process: Streamlined and optimized
- ✅ Test Coverage: Comprehensive with parallel execution
- ✅ Performance: 71% faster parallel execution vs sequential
- ✅ Stability: Full backward compatibility maintained
We welcome contributions! Here's how to get started:
# Clone the repository
git clone https://github.com/ruvnet/claude-code-flow.git
cd claude-code-flow
# Install dependencies and setup
deno cache --reload src/deps.ts
# Run tests
deno task test
# Build the project
deno task build- 🐛 Bug Reports: Use GitHub issues with detailed reproduction steps
- 💡 Feature Requests: Propose new features with use cases
- 🔧 Pull Requests: Follow our coding standards and include tests
- 📚 Documentation: Help improve docs and examples
MIT License - see LICENSE for details.
- Anthropic: For the amazing Claude AI that powers this platform
- Deno Team: For the excellent TypeScript runtime
- Open Source Community: For contributions and feedback
- SPARC Methodology: For the structured development approach
npx claude-flow@latest init --sparcJoin thousands of developers already using Claude-Flow!
Built with ❤️ by rUv | Powered by Claude AI