# Claude MPM - Multi-Agent Project Manager
> **Note**: This project is a fork of [claude-multiagent-pm](https://github.com/kfsone/claude-multiagent-pm), enhanced to integrate with [Claude Code](https://docs.anthropic.com/en/docs/claude-code) v1.0.60+'s native agent system. This integration enables seamless orchestration of Claude Code's built-in agents (research, engineer, qa, documentation, security, ops, version_control, data_engineer) through a unified project management interface.
> **⚠️ Version 2.0.0 Breaking Changes**: Agent schema has been standardized. Agent IDs no longer use the `_agent` suffix (e.g., `research_agent` → `research`). See the [migration guide](docs/user/05-migration/schema-standardization-migration.md) for details.
A framework for Claude that enables multi-agent workflows and extensible capabilities through a modular architecture.
## Quick Start
### Why Interactive Mode?
**Interactive mode is significantly more performant** than non-interactive commands. It maintains context between requests and avoids the overhead of repeatedly launching Claude, making your development workflow much faster and more efficient.
### Installation
```bash
# Install globally via npm (recommended)
npm install -g @bobmatnyc/claude-mpm
# Or use npx for one-time usage
npx @bobmatnyc/claude-mpm
```
### Three Essential Use Cases
#### 1. 🔍 **Understand Your Codebase**
Start with codebase exploration - perfect for onboarding or getting oriented:
```bash
# Launch interactive mode
claude-mpm
# Then ask:
> Explain the codebase structure. What are the main components, how do they interact, and what architectural patterns are used?
```
#### 2. 🚀 **Build a New Project**
For greenfield development, use detailed, AI-generated prompts for best results:
```bash
claude-mpm
# Example detailed prompt (AI-generated prompts work best):
> Create a modern web application with the following requirements:
> - Next.js 14 with TypeScript and Tailwind CSS
> - Authentication using NextAuth.js with GitHub provider
> - PostgreSQL database with Prisma ORM
> - User dashboard with CRUD operations for "projects"
> - API routes following REST conventions
> - Responsive design with dark/light mode toggle
> - Form validation using react-hook-form and zod
> - Include proper error handling and loading states
> - Set up ESLint, Prettier, and basic testing with Jest
> - Generate a complete project structure with all necessary files
```
#### 3. 🔧 **Enhance Existing Code**
For working on your current codebase, provide rich context:
```bash
claude-mpm
# Example detailed enhancement prompt:
> I need to add real-time notifications to my existing Next.js application. Current tech stack:
> - Next.js 13 with app router
> - TypeScript
> - Tailwind CSS
> - PostgreSQL with Prisma
> - User authentication already implemented
>
> Requirements:
> - WebSocket-based real-time notifications
> - Toast notifications in the UI
> - Database table to store notification history
> - Mark as read/unread functionality
> - Different notification types (info, warning, success, error)
> - Admin panel to send system-wide notifications
> - Email fallback for offline users
>
> Please analyze my current codebase structure and implement this feature following my existing patterns and conventions.
```
### 💡 Pro Tips for Better Results
1. **Use AI to generate your prompts**: Ask ChatGPT or Claude to help you create detailed, specific prompts for your use case
2. **Provide context**: Include your tech stack, requirements, and any constraints
3. **Stay interactive**: Keep the conversation going to refine and iterate on solutions
4. **Ask for explanations**: Request explanations of architectural decisions and trade-offs
### Alternative: Non-Interactive Mode
For automation or simple one-off tasks:
```bash
# Quick analysis
claude-mpm run -i "What testing frameworks are used in this project?" --non-interactive
# With subprocess orchestration for complex tasks
claude-mpm run --subprocess -i "Audit this codebase for security vulnerabilities" --non-interactive
```
**Note**: Non-interactive mode has higher overhead and is less efficient for multi-step development workflows.
## 📚 Documentation
- **[User Guide](docs/user/)** - Getting started, usage, and troubleshooting
- **[Developer Guide](docs/developer/)** - Architecture, API reference, and contributing
- **[Design Documents](docs/design/)** - Architectural decisions and design patterns
- **[Differences from claude-multiagent-pm](docs/user/differences-from-claude-multiagent-pm.md)** - Migration guide
## Why Claude MPM?
Claude MPM provides a modular framework for extending Claude's capabilities:
- **🧩 Modular Architecture**: Extensible agent system and hook-based customization
- **🤖 Multi-Agent Support**: Specialized agents for different tasks
- **📝 Comprehensive Logging**: Every interaction is logged for review
- **🛠️ Service-Based Design**: Clean separation of concerns through services
## How It Works
```
┌─────────────┐ ┌─────────────────┐ ┌──────────────┐
│ Terminal │──────▶│ Claude MPM │──────▶│ Services │
│ (User) │ │ Framework │ │ & Agents │
└─────────────┘ └─────────────────┘ └──────────────┘
│
┌────────┴────────┐
│ │
┌─────▼─────┐ ┌─────▼─────┐
│ Hook │ │ Agent │
│ System │ │ Registry │
└───────────┘ └───────────┘
```
## Overview
Claude MPM provides a modular framework for extending Claude's capabilities:
- **Agent System**: Specialized agents for different task types
- **Hook System**: Extensible architecture through pre/post hooks
- **Service Layer**: Clean separation of business logic
- **Agent Registry**: Dynamic agent discovery and loading
- **Session Logging**: Comprehensive logging of all interactions
## Key Features
### Security Features
- **File Access Restrictions**: Agents are restricted to their working directory
- **Bash Command Validation**: Comprehensive validation prevents writes outside project directory
- **Path Traversal Protection**: Blocks attempts to escape the working directory
- **Dangerous Command Blocking**: Prevents execution of sudo, rm -rf, and other risky commands
- **Security Audit Logging**: All file operations are logged for security monitoring
- See [Security Documentation](docs/SECURITY.md) for details
### Agent System
- Specialized agents for different domains (Research, Engineer, etc.)
- Dynamic agent discovery and registration
- Template-based agent definitions
- Extensible agent architecture
- **Dynamic Capabilities**: Agent documentation automatically generated from deployed agents
### Hook System
- Pre and post-processing hooks
- Customizable behavior injection
- Plugin-like extensibility
- Clean integration points
### Service Architecture
- Modular service components
- Clean separation of concerns
- Reusable business logic
- Well-defined interfaces
### Session Management
- Comprehensive logging of all interactions
- Debug mode for troubleshooting
- Organized log structure
- Performance monitoring
### Security Features
- **File System Protection**: Automatic sandboxing prevents file operations outside the working directory
- **Path Traversal Prevention**: Blocks attempts to escape the project directory using `..` or symlinks
- **Write Operation Control**: All write operations are validated while read operations remain unrestricted
- **Agent-Level Restrictions**: Each agent can have custom file access boundaries via `file_access` configuration
- **PM Agent Orchestration**: New PM (Project Manager) agent ensures all sub-agents operate within security boundaries
- **Transparent Security**: Zero-configuration security that works automatically in the background
- **Comprehensive Logging**: All security events are logged for audit purposes
## Installation
### Other Installation Methods
#### Using UV (Recommended - Fast)
UV is a lightning-fast Python package manager written in Rust, offering 10-100x speed improvements over pip.
```bash
# Install UV (if not already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh
# Install claude-mpm with UV
uv pip install claude-mpm
# Or install from git
uv pip install git+https://github.com/bobmatnyc/claude-mpm.git
```
#### Using pip (Traditional)
```bash
# Install from PyPI
pip install claude-mpm
# Or install from git
pip install git+https://github.com/bobmatnyc/claude-mpm.git
```
### From Source (Development)
```bash
# Clone the repository
git clone https://github.com/bobmatnyc/claude-mpm.git
cd claude-mpm
# Option A: Using UV (Recommended - Much faster)
uv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
uv pip install -e .
# Option B: Traditional approach
./install_dev.sh
source venv/bin/activate
```
### Dependencies
#### Core Requirements
- Python 3.8+
- Claude Code CLI 1.0.60+ (must be installed and in PATH)
#### Automatically Installed
- tree-sitter & language packs (for code analysis)
- All other Python dependencies
#### Code Analysis Dependencies
- **tree-sitter** (>=0.21.0) - Core parsing library for advanced code analysis
- **tree-sitter-language-pack** (>=0.20.0) - Multi-language support package providing parsers for 41+ programming languages
These tree-sitter dependencies enable:
- **Advanced Code Analysis**: The TreeSitterAnalyzer component provides syntax-aware code parsing for Research Agent operations
- **Agent Modification Tracking**: Real-time analysis of agent code changes with AST-level understanding
- **Multi-Language Support**: Out-of-the-box support for Python, JavaScript, TypeScript, Go, Rust, Java, C/C++, and 35+ other languages
- **Performance**: Fast, incremental parsing suitable for real-time analysis during agent operations
## Usage
### Basic Usage
```bash
# Run interactive session
claude-mpm
# Run with debug logging
claude-mpm --debug
# Show configuration info
claude-mpm info
```
### Command Line Options
```
claude-mpm [-h] [-d] [--log-dir LOG_DIR] {run,info}
Options:
-d, --debug Enable debug logging
--log-dir LOG_DIR Custom log directory (default: ~/.claude-mpm/logs)
Commands:
run Run Claude session (default)
info Show framework and configuration info
```
## Architecture
### Core Components
```
claude-mpm/
├── src/claude_mpm/
│ ├── agents/ # Agent templates
│ ├── core/ # Core functionality
│ │ ├── agent_registry.py # Agent discovery
│ │ └── simple_runner.py # Main runner
│ ├── services/ # Business logic
│ │ ├── hook_service.py
│ │ └── agent_deployment.py
│ ├── hooks/ # Hook system
│ └── cli/ # CLI interface
└── docs/ # Organized documentation
├── user/ # User guides
├── developer/ # Developer docs
└── design/ # Architecture docs
```
## Testing
```bash
# Run all tests
./scripts/run_all_tests.sh
# Run E2E tests
./scripts/run_e2e_tests.sh
# Run specific test
pytest tests/test_orchestrator.py -v
```
## Logging
Logs are stored in `~/.claude-mpm/logs/` by default:
- `mpm_YYYYMMDD_HHMMSS.log` - Detailed debug logs
- `latest.log` - Symlink to most recent log
- Session logs in `~/.claude-mpm/sessions/`
## Development
For detailed development information, see the [Developer Documentation](docs/developer/).
### Quick Start
```bash
# Install development dependencies
pip install -e ".[dev]"
# Run tests
python run_tests.py
# Test agent integration
python examples/test_agent_integration.py
```
### Key Resources
- [Architecture Overview](docs/developer/README.md#architecture-overview)
- [Project Structure](docs/developer/STRUCTURE.md)
- [Testing Guide](docs/developer/QA.md)
- [API Reference](docs/developer/README.md#api-reference)
- [Contributing Guide](docs/developer/README.md#contributing)
## Troubleshooting
For detailed troubleshooting, see the [User Guide](docs/user/README.md#troubleshooting).
### Quick Fixes
**Claude not found**
```bash
which claude # Check if Claude is in PATH
```
**Debug mode**
```bash
claude-mpm --debug # Enable debug logging
tail -f ~/.claude-mpm/logs/latest.log # View logs
```
## License
MIT License - See LICENSE file for details
Raw data
{
"_id": null,
"home_page": "https://github.com/bobmatnyc/claude-mpm",
"name": "claude-mpm",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "claude, orchestration, multi-agent, ticket-management",
"author": "Claude MPM Team",
"author_email": "bob@matsuoka.com",
"download_url": "https://files.pythonhosted.org/packages/a3/a6/c5ec48c6c41ad9fa6a75b9df2ee56f16564df9a4bdba43b631e3ddb5cdec/claude_mpm-3.1.3.tar.gz",
"platform": null,
"description": "# Claude MPM - Multi-Agent Project Manager\n\n> **Note**: This project is a fork of [claude-multiagent-pm](https://github.com/kfsone/claude-multiagent-pm), enhanced to integrate with [Claude Code](https://docs.anthropic.com/en/docs/claude-code) v1.0.60+'s native agent system. This integration enables seamless orchestration of Claude Code's built-in agents (research, engineer, qa, documentation, security, ops, version_control, data_engineer) through a unified project management interface.\n\n> **\u26a0\ufe0f Version 2.0.0 Breaking Changes**: Agent schema has been standardized. Agent IDs no longer use the `_agent` suffix (e.g., `research_agent` \u2192 `research`). See the [migration guide](docs/user/05-migration/schema-standardization-migration.md) for details.\n\nA framework for Claude that enables multi-agent workflows and extensible capabilities through a modular architecture.\n\n## Quick Start\n\n### Why Interactive Mode?\n**Interactive mode is significantly more performant** than non-interactive commands. It maintains context between requests and avoids the overhead of repeatedly launching Claude, making your development workflow much faster and more efficient.\n\n### Installation\n\n```bash\n# Install globally via npm (recommended)\nnpm install -g @bobmatnyc/claude-mpm\n\n# Or use npx for one-time usage\nnpx @bobmatnyc/claude-mpm\n```\n\n### Three Essential Use Cases\n\n#### 1. \ud83d\udd0d **Understand Your Codebase**\nStart with codebase exploration - perfect for onboarding or getting oriented:\n\n```bash\n# Launch interactive mode\nclaude-mpm\n\n# Then ask:\n> Explain the codebase structure. What are the main components, how do they interact, and what architectural patterns are used?\n```\n\n#### 2. \ud83d\ude80 **Build a New Project** \nFor greenfield development, use detailed, AI-generated prompts for best results:\n\n```bash\nclaude-mpm\n\n# Example detailed prompt (AI-generated prompts work best):\n> Create a modern web application with the following requirements:\n> - Next.js 14 with TypeScript and Tailwind CSS\n> - Authentication using NextAuth.js with GitHub provider\n> - PostgreSQL database with Prisma ORM\n> - User dashboard with CRUD operations for \"projects\"\n> - API routes following REST conventions\n> - Responsive design with dark/light mode toggle\n> - Form validation using react-hook-form and zod\n> - Include proper error handling and loading states\n> - Set up ESLint, Prettier, and basic testing with Jest\n> - Generate a complete project structure with all necessary files\n```\n\n#### 3. \ud83d\udd27 **Enhance Existing Code**\nFor working on your current codebase, provide rich context:\n\n```bash\nclaude-mpm\n\n# Example detailed enhancement prompt:\n> I need to add real-time notifications to my existing Next.js application. Current tech stack:\n> - Next.js 13 with app router\n> - TypeScript\n> - Tailwind CSS\n> - PostgreSQL with Prisma\n> - User authentication already implemented\n> \n> Requirements:\n> - WebSocket-based real-time notifications\n> - Toast notifications in the UI\n> - Database table to store notification history\n> - Mark as read/unread functionality\n> - Different notification types (info, warning, success, error)\n> - Admin panel to send system-wide notifications\n> - Email fallback for offline users\n> \n> Please analyze my current codebase structure and implement this feature following my existing patterns and conventions.\n```\n\n### \ud83d\udca1 Pro Tips for Better Results\n\n1. **Use AI to generate your prompts**: Ask ChatGPT or Claude to help you create detailed, specific prompts for your use case\n2. **Provide context**: Include your tech stack, requirements, and any constraints\n3. **Stay interactive**: Keep the conversation going to refine and iterate on solutions\n4. **Ask for explanations**: Request explanations of architectural decisions and trade-offs\n\n### Alternative: Non-Interactive Mode\nFor automation or simple one-off tasks:\n\n```bash\n# Quick analysis\nclaude-mpm run -i \"What testing frameworks are used in this project?\" --non-interactive\n\n# With subprocess orchestration for complex tasks\nclaude-mpm run --subprocess -i \"Audit this codebase for security vulnerabilities\" --non-interactive\n```\n\n**Note**: Non-interactive mode has higher overhead and is less efficient for multi-step development workflows.\n\n\n## \ud83d\udcda Documentation\n\n- **[User Guide](docs/user/)** - Getting started, usage, and troubleshooting\n- **[Developer Guide](docs/developer/)** - Architecture, API reference, and contributing\n- **[Design Documents](docs/design/)** - Architectural decisions and design patterns\n- **[Differences from claude-multiagent-pm](docs/user/differences-from-claude-multiagent-pm.md)** - Migration guide\n\n## Why Claude MPM?\n\nClaude MPM provides a modular framework for extending Claude's capabilities:\n\n- **\ud83e\udde9 Modular Architecture**: Extensible agent system and hook-based customization\n- **\ud83e\udd16 Multi-Agent Support**: Specialized agents for different tasks\n- **\ud83d\udcdd Comprehensive Logging**: Every interaction is logged for review\n- **\ud83d\udee0\ufe0f Service-Based Design**: Clean separation of concerns through services\n\n## How It Works\n\n```\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n\u2502 Terminal \u2502\u2500\u2500\u2500\u2500\u2500\u2500\u25b6\u2502 Claude MPM \u2502\u2500\u2500\u2500\u2500\u2500\u2500\u25b6\u2502 Services \u2502\n\u2502 (User) \u2502 \u2502 Framework \u2502 \u2502 & Agents \u2502\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2534\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u25bc\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u25bc\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Hook \u2502 \u2502 Agent \u2502\n \u2502 System \u2502 \u2502 Registry \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n```\n\n## Overview\n\nClaude MPM provides a modular framework for extending Claude's capabilities:\n\n- **Agent System**: Specialized agents for different task types\n- **Hook System**: Extensible architecture through pre/post hooks\n- **Service Layer**: Clean separation of business logic\n- **Agent Registry**: Dynamic agent discovery and loading\n- **Session Logging**: Comprehensive logging of all interactions\n\n## Key Features\n\n### Security Features\n- **File Access Restrictions**: Agents are restricted to their working directory\n- **Bash Command Validation**: Comprehensive validation prevents writes outside project directory\n- **Path Traversal Protection**: Blocks attempts to escape the working directory\n- **Dangerous Command Blocking**: Prevents execution of sudo, rm -rf, and other risky commands\n- **Security Audit Logging**: All file operations are logged for security monitoring\n- See [Security Documentation](docs/SECURITY.md) for details\n\n### Agent System\n- Specialized agents for different domains (Research, Engineer, etc.)\n- Dynamic agent discovery and registration\n- Template-based agent definitions\n- Extensible agent architecture\n- **Dynamic Capabilities**: Agent documentation automatically generated from deployed agents\n\n### Hook System\n- Pre and post-processing hooks\n- Customizable behavior injection\n- Plugin-like extensibility\n- Clean integration points\n\n### Service Architecture\n- Modular service components\n- Clean separation of concerns\n- Reusable business logic\n- Well-defined interfaces\n\n### Session Management\n- Comprehensive logging of all interactions\n- Debug mode for troubleshooting\n- Organized log structure\n- Performance monitoring\n\n### Security Features\n- **File System Protection**: Automatic sandboxing prevents file operations outside the working directory\n- **Path Traversal Prevention**: Blocks attempts to escape the project directory using `..` or symlinks\n- **Write Operation Control**: All write operations are validated while read operations remain unrestricted\n- **Agent-Level Restrictions**: Each agent can have custom file access boundaries via `file_access` configuration\n- **PM Agent Orchestration**: New PM (Project Manager) agent ensures all sub-agents operate within security boundaries\n- **Transparent Security**: Zero-configuration security that works automatically in the background\n- **Comprehensive Logging**: All security events are logged for audit purposes\n\n## Installation\n\n### Other Installation Methods\n\n#### Using UV (Recommended - Fast)\nUV is a lightning-fast Python package manager written in Rust, offering 10-100x speed improvements over pip.\n\n```bash\n# Install UV (if not already installed)\ncurl -LsSf https://astral.sh/uv/install.sh | sh\n\n# Install claude-mpm with UV\nuv pip install claude-mpm\n\n# Or install from git\nuv pip install git+https://github.com/bobmatnyc/claude-mpm.git\n```\n\n#### Using pip (Traditional)\n```bash\n# Install from PyPI\npip install claude-mpm\n\n# Or install from git\npip install git+https://github.com/bobmatnyc/claude-mpm.git\n```\n\n### From Source (Development)\n\n```bash\n# Clone the repository\ngit clone https://github.com/bobmatnyc/claude-mpm.git\ncd claude-mpm\n\n# Option A: Using UV (Recommended - Much faster)\nuv venv\nsource venv/bin/activate # On Windows: venv\\Scripts\\activate\nuv pip install -e .\n\n# Option B: Traditional approach\n./install_dev.sh\nsource venv/bin/activate\n```\n\n### Dependencies\n\n#### Core Requirements\n- Python 3.8+\n- Claude Code CLI 1.0.60+ (must be installed and in PATH)\n\n#### Automatically Installed\n- tree-sitter & language packs (for code analysis)\n- All other Python dependencies\n\n#### Code Analysis Dependencies\n- **tree-sitter** (>=0.21.0) - Core parsing library for advanced code analysis\n- **tree-sitter-language-pack** (>=0.20.0) - Multi-language support package providing parsers for 41+ programming languages\n\nThese tree-sitter dependencies enable:\n- **Advanced Code Analysis**: The TreeSitterAnalyzer component provides syntax-aware code parsing for Research Agent operations\n- **Agent Modification Tracking**: Real-time analysis of agent code changes with AST-level understanding\n- **Multi-Language Support**: Out-of-the-box support for Python, JavaScript, TypeScript, Go, Rust, Java, C/C++, and 35+ other languages\n- **Performance**: Fast, incremental parsing suitable for real-time analysis during agent operations\n\n## Usage\n\n### Basic Usage\n\n```bash\n# Run interactive session\nclaude-mpm\n\n# Run with debug logging\nclaude-mpm --debug\n\n# Show configuration info\nclaude-mpm info\n```\n\n\n### Command Line Options\n\n```\nclaude-mpm [-h] [-d] [--log-dir LOG_DIR] {run,info}\n\nOptions:\n -d, --debug Enable debug logging\n --log-dir LOG_DIR Custom log directory (default: ~/.claude-mpm/logs)\n\nCommands:\n run Run Claude session (default)\n info Show framework and configuration info\n```\n\n\n\n## Architecture\n\n### Core Components\n\n```\nclaude-mpm/\n\u251c\u2500\u2500 src/claude_mpm/\n\u2502 \u251c\u2500\u2500 agents/ # Agent templates\n\u2502 \u251c\u2500\u2500 core/ # Core functionality\n\u2502 \u2502 \u251c\u2500\u2500 agent_registry.py # Agent discovery\n\u2502 \u2502 \u2514\u2500\u2500 simple_runner.py # Main runner\n\u2502 \u251c\u2500\u2500 services/ # Business logic\n\u2502 \u2502 \u251c\u2500\u2500 hook_service.py\n\u2502 \u2502 \u2514\u2500\u2500 agent_deployment.py\n\u2502 \u251c\u2500\u2500 hooks/ # Hook system\n\u2502 \u2514\u2500\u2500 cli/ # CLI interface\n\u2514\u2500\u2500 docs/ # Organized documentation\n \u251c\u2500\u2500 user/ # User guides\n \u251c\u2500\u2500 developer/ # Developer docs\n \u2514\u2500\u2500 design/ # Architecture docs\n```\n\n## Testing\n\n```bash\n# Run all tests\n./scripts/run_all_tests.sh\n\n# Run E2E tests\n./scripts/run_e2e_tests.sh\n\n# Run specific test\npytest tests/test_orchestrator.py -v\n```\n\n## Logging\n\nLogs are stored in `~/.claude-mpm/logs/` by default:\n\n- `mpm_YYYYMMDD_HHMMSS.log` - Detailed debug logs\n- `latest.log` - Symlink to most recent log\n- Session logs in `~/.claude-mpm/sessions/`\n\n\n## Development\n\nFor detailed development information, see the [Developer Documentation](docs/developer/).\n\n### Quick Start\n\n```bash\n# Install development dependencies\npip install -e \".[dev]\"\n\n# Run tests\npython run_tests.py\n\n# Test agent integration\npython examples/test_agent_integration.py\n```\n\n### Key Resources\n\n- [Architecture Overview](docs/developer/README.md#architecture-overview)\n- [Project Structure](docs/developer/STRUCTURE.md)\n- [Testing Guide](docs/developer/QA.md)\n- [API Reference](docs/developer/README.md#api-reference)\n- [Contributing Guide](docs/developer/README.md#contributing)\n\n## Troubleshooting\n\nFor detailed troubleshooting, see the [User Guide](docs/user/README.md#troubleshooting).\n\n### Quick Fixes\n\n**Claude not found**\n```bash\nwhich claude # Check if Claude is in PATH\n```\n\n\n**Debug mode**\n```bash\nclaude-mpm --debug # Enable debug logging\ntail -f ~/.claude-mpm/logs/latest.log # View logs\n```\n\n## License\n\nMIT License - See LICENSE file for details\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Claude Multi-agent Project Manager - Clean orchestration with ticket management",
"version": "3.1.3",
"project_urls": {
"Homepage": "https://github.com/bobmatnyc/claude-mpm"
},
"split_keywords": [
"claude",
" orchestration",
" multi-agent",
" ticket-management"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "6744a8f3503854cb49902a43737c6169f408f6f203f3efa2b08a6126b1249127",
"md5": "bcc67de97c316a457c560b045c68790e",
"sha256": "f776c5d6eac4efae55fda8867f24635e005ab559a53938a73f606cbd2b6a99f2"
},
"downloads": -1,
"filename": "claude_mpm-3.1.3-py3-none-any.whl",
"has_sig": false,
"md5_digest": "bcc67de97c316a457c560b045c68790e",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 509635,
"upload_time": "2025-07-29T21:20:06",
"upload_time_iso_8601": "2025-07-29T21:20:06.783155Z",
"url": "https://files.pythonhosted.org/packages/67/44/a8f3503854cb49902a43737c6169f408f6f203f3efa2b08a6126b1249127/claude_mpm-3.1.3-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "a3a6c5ec48c6c41ad9fa6a75b9df2ee56f16564df9a4bdba43b631e3ddb5cdec",
"md5": "67e68893cac3648a47a938296ae8833e",
"sha256": "5ce531d7b1de91e6305f51faef6bd368e1fe76df42cf3a32bbb9324423c93aeb"
},
"downloads": -1,
"filename": "claude_mpm-3.1.3.tar.gz",
"has_sig": false,
"md5_digest": "67e68893cac3648a47a938296ae8833e",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 1896131,
"upload_time": "2025-07-29T21:20:08",
"upload_time_iso_8601": "2025-07-29T21:20:08.474848Z",
"url": "https://files.pythonhosted.org/packages/a3/a6/c5ec48c6c41ad9fa6a75b9df2ee56f16564df9a4bdba43b631e3ddb5cdec/claude_mpm-3.1.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-07-29 21:20:08",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "bobmatnyc",
"github_project": "claude-mpm",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "ai-trackdown-pytools",
"specs": [
[
">=",
"1.4.0"
]
]
},
{
"name": "pyyaml",
"specs": [
[
">=",
"6.0"
]
]
},
{
"name": "python-dotenv",
"specs": [
[
">=",
"0.19.0"
]
]
},
{
"name": "rich",
"specs": [
[
">=",
"13.0.0"
]
]
},
{
"name": "click",
"specs": [
[
">=",
"8.0.0"
]
]
},
{
"name": "pexpect",
"specs": [
[
">=",
"4.8.0"
]
]
},
{
"name": "psutil",
"specs": [
[
">=",
"5.9.0"
]
]
},
{
"name": "requests",
"specs": [
[
">=",
"2.25.0"
]
]
},
{
"name": "flask",
"specs": [
[
">=",
"3.0.0"
]
]
},
{
"name": "flask-cors",
"specs": [
[
">=",
"4.0.0"
]
]
},
{
"name": "watchdog",
"specs": [
[
">=",
"3.0.0"
]
]
},
{
"name": "tree-sitter",
"specs": [
[
">=",
"0.21.0"
]
]
},
{
"name": "tree-sitter-language-pack",
"specs": [
[
">=",
"0.20.0"
]
]
}
],
"lcname": "claude-mpm"
}
JSON
