Open-source Platform to Build Your Own Personal AI Assistant, mashup of clawdbot and n8n, but More Powerful.
60 nodes | 6 AI providers | WebSocket-first | Self-hosted
npm install -g machinaos
machinaos startOther Install Options
Linux/macOS:
curl -fsSL https://raw.githubusercontent.com/trohitg/MachinaOS/main/install.sh | bashWindows (PowerShell):
iwr -useb https://raw.githubusercontent.com/trohitg/MachinaOS/main/install.ps1 | iexgit clone https://github.com/trohitg/MachinaOS.git
cd MachinaOS
npm run build
npm run startgit clone https://github.com/trohitg/MachinaOS.git
cd MachinaOS
npm run docker:up| Provider | Models | Features |
|---|---|---|
| OpenAI | GPT-4o, GPT-4 Turbo, o1, o3, o4-mini | JSON mode, reasoning effort |
| Anthropic | Claude 3.5 Sonnet, Claude 3 Opus, Claude 3 Haiku | Extended thinking |
| Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.0 Flash Thinking | Multimodal, 1M context | |
| OpenRouter | 200+ models | Unified API for all providers |
| Groq | Llama, Mixtral, Qwen | Ultra-fast inference |
| Cerebras | Llama, Qwen | Ultra-fast on custom hardware |
- AI Agent - LangGraph-powered with tool calling and iterative reasoning
- Chat Agent - Conversational agent with skill support for multi-turn chat
- 10 Skills - WhatsApp, Maps, HTTP, Scheduler, Android, Code, Memory, Web Search, Custom
- 4 Tools - Calculator, Current Time, Web Search, Android Toolkit
- Simple Memory - Markdown-based conversation history with vector storage
- WhatsApp - Send/receive messages with QR pairing, filters, group support
- Android - 17 service nodes for device control (battery, WiFi, Bluetooth, apps, camera, sensors)
- HTTP/Webhooks - REST API integration with event-driven triggers
- Google Maps - Geocoding, nearby places, directions
- HTTP Scraper - Scrape URLs with pagination and date ranges
- File Downloader - Parallel downloads with semaphore concurrency
- Document Parser - PyPDF, Marker (OCR), Unstructured, BeautifulSoup
- Text Chunker - Recursive, markdown, or token-based splitting
- Embedding Generator - HuggingFace, OpenAI, Ollama embeddings
- Vector Store - ChromaDB, Qdrant, Pinecone backends
| Category | Count | Description |
|---|---|---|
| AI Models | 6 | OpenAI, Anthropic, Google, OpenRouter, Groq, Cerebras |
| AI Agents | 3 | AI Agent, Chat Agent, Simple Memory |
| AI Skills | 10 | WhatsApp, Maps, HTTP, Scheduler, Android, Code, etc. |
| AI Tools | 4 | Calculator, Time, Search, Android Toolkit |
| 3 | Send, Receive, Database | |
| Android | 17 | Device control and monitoring |
| Documents | 6 | RAG pipeline nodes |
| Utilities | 5 | HTTP, Webhooks, Chat Trigger, Console |
| Location | 3 | Google Maps integration |
| Code | 2 | Python and JavaScript executors |
| Workflow | 1 | Start node |
Total: 60 nodes
The install script handles these automatically, but for manual installation:
- Node.js 18+ - https://nodejs.org/
- Python 3.11+ - https://python.org/
- uv -
curl -LsSf https://astral.sh/uv/install.sh | sh - Go 1.21+ - https://go.dev/dl/ (for WhatsApp service)
| Command | Description |
|---|---|
machinaos start |
Start all services (frontend, backend, WhatsApp) |
machinaos stop |
Stop all running services |
machinaos build |
Build for production |
machinaos clean |
Remove build artifacts and dependencies |
machinaos docker:up |
Start with Docker Compose |
machinaos docker:down |
Stop Docker containers |
machinaos help |
Show all available commands |
API Keys: Click Credentials button in toolbar to add API keys for OpenAI, Claude, Google Maps, etc.
Environment: Copy .env.template to .env and customize ports, auth settings, database location.
| Command | Description |
|---|---|
machinaos docker:up |
Start containers (detached) |
machinaos docker:down |
Stop containers |
machinaos docker:build |
Rebuild images |
machinaos docker:logs |
View logs (follows) |
Redis (optional): Set REDIS_ENABLED=true in .env
npm run docker:prod:build
npm run docker:prod:upMachinaOS/
├── client/ # React frontend (localhost:3000)
├── server/ # Python FastAPI backend (localhost:3010)
│ ├── services/ # Workflow execution, AI, handlers
│ ├── routers/ # API endpoints, WebSocket
│ └── whatsapp-rpc/ # WhatsApp Go service
├── scripts/ # Cross-platform Node.js scripts
└── bin/cli.js # CLI entry point
- Frontend: React 19, TypeScript, React Flow, Zustand
- Backend: Python 3.11+, FastAPI, SQLite, LangChain/LangGraph
- Services: WhatsApp (Go + whatsmeow), WebSocket relay
- Package Manager: uv (Python), npm (Node.js)
Full documentation available at: https://docs.machinaos.dev
Port already in use:
machinaos stop # Kill all services
machinaos start # RestartMissing dependencies:
machinaos build # Install all dependenciesClean install:
machinaos clean # Remove node_modules, .venv, dist
machinaos build # Reinstall everything- Fork the repository
- Create a feature branch
- Make your changes
- Run tests and linting
- Submit a pull request
MIT