Open-source personal AI assistant and Agent Harness that plans tasks, runs tools and skills, and keeps memory across web and IM channels.
spinabot/brigade
A self-hosted personal AI agent platform where a crew of isolated agents shares long-term memory and reaches you through terminal, WhatsApp, Telegram, Slack and more.
Open-source alternative to:ChatGPT
- Stars
- 11,353
- Forks
- 76
- Open issues
- 38
- Last commit
- 6 days ago
What is it?
Brigade is a self-hosted personal AI assistant framework built around a crew of agents. Each agent has its own workspace, persona, credentials and memory, and agents can be arranged into an org chart that controls which agents may message each other. A shared long-term memory engine called Tideline stores facts with origin scoping and decay, and supports hybrid keyword and vector recall.
The gateway is a long-running process that holds all state, runs agent turns and cron jobs, and accepts connections from a terminal chat UI, the CLI, and messaging channels such as WhatsApp, Telegram, Slack, Discord, iMessage and BlueBubbles. It works with models from Anthropic, OpenAI, Gemini, OpenRouter, Ollama and other OpenAI-compatible endpoints, and includes sub-agent delegation, a skill system, scheduled jobs, and connectors to over 1,000 apps through Composio. Data lives under ~/.brigade/ by default in filesystem mode, with an optional self-hosted Convex backend.
The project is written in TypeScript, installs from npm or a shell script, and includes an MCP server that exposes its memory to other MCP clients.
Key features
- 01Chat TUI and CLI connected to a long-running gateway that keeps agents and cron jobs alive
- 02Isolated agents with their own workspaces, credentials, memory and org-chart communication rules
- 03Messaging channels including WhatsApp, Telegram, Slack, Discord, iMessage and BlueBubbles
- 04Long-term memory engine with hybrid keyword and vector recall, exposed as an MCP server
- 05Connectors to 1,000+ apps via Composio, plus document reading and creation for Office and PDF files
- 06Works with Claude, GPT, Gemini, OpenRouter and local Ollama models, switchable mid-conversation
Use cases
- Running a personal assistant on a home server or Raspberry Pi that answers over WhatsApp or Telegram
- Giving several specialized agents separate memory and permissions while letting them delegate tasks to each other
- Scheduling recurring agent tasks such as daily summaries or reminders with cron expressions
- Connecting an AI assistant to Gmail, Slack, GitHub or Notion through a single self-hosted gateway
Quick start
- 1
Install from npm (requires Node 22.12+)
npm i -g @spinabot/brigade - 2
Run the onboarding wizard to pick a provider and model
brigade onboard - 3
Start the always-on gateway
brigade gateway run - 4
Chat with your crew in the terminal
brigade tui
Things to know
- Requires Node 22.12 or newer to install from npm.
- Using a model needs an API key or a supported subscription such as Claude, ChatGPT, or Copilot.
Topics
FAQ
- What is brigade?
- A self-hosted personal AI agent platform where a crew of isolated agents shares long-term memory and reaches you through terminal, WhatsApp, Telegram, Slack and more. Brigade is a self-hosted personal AI assistant framework built around a crew of agents. Each agent has its own workspace, persona, credentials and memory, and agents can be arranged into an org chart that controls which agents may message each other. A shared long-term memory engine called Tideline stores facts with origin scoping and decay, and supports hybrid keyword and vector recall.
- Is brigade free to use? What license does it have?
- brigade is open source under the MIT license. Check the license terms before using it in commercial products.
- Is brigade still maintained?
- Actively maintained. The last commit to brigade was on Oct 16, 2026.
- What can I use brigade for?
- Running a personal assistant on a home server or Raspberry Pi that answers over WhatsApp or Telegram Giving several specialized agents separate memory and permissions while letting them delegate tasks to each other Scheduling recurring agent tasks such as daily summaries or reminders with cron expressions Connecting an AI assistant to Gmail, Slack, GitHub or Notion through a single self-hosted gateway
- How do I install brigade?
- Install from npm (requires Node 22.12+): npm i -g @spinabot/brigade Run the onboarding wizard to pick a provider and model: brigade onboard Start the always-on gateway: brigade gateway run Chat with your crew in the terminal: brigade tui
- Where can I get brigade?
- The source code and installation instructions are on GitHub: https://github.com/spinabot/brigade
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Self-hosted personal AI agent framework in Python with a browser WebUI, terminal client, chat app channels, tools, memory and MCP support.
A self-improving AI agent that creates and refines skills from experience, runs on many model providers, and works through CLI and messaging apps.
A token-efficient, microkernel AI agent for CLI, Web UI and chat channels that routes each turn to the cheapest capable model.
A local-first desktop AI assistant with long-term memory, autonomous goals, dynamic workflows, and a headless server mode for NAS or cloud hosts.
Self-hostable open-source platform for running AI agents on any model and any machine, with tools, permissions and streaming built in.