AI agents that know you and your work, in a chat app anyone can use.
Open source under Apache 2.0 · Any model, local or cloud · Self-host the whole stack
Website · Docs · Showcase · MindRoom Chat · Hosted
film-readme-github-dark.mp4
MindRoom gives you a personal agent for your calendar, notes, trips, and homelab, and shared agents for your team's email, documents, and code. Every agent is a real user on Matrix, the open chat standard, so you talk to it in MindRoom Chat, in any other Matrix client, or in Slack, Telegram, WhatsApp, and Discord through bridges. Pick a local model for your private life or a frontier model for hard problems, and self-host the whole stack or run only the agents on your own computer.
Run it on any computer
uvx mindroom runNeeds uv and a model: an API key, a subscription login such as Codex, or a local model. It installs MindRoom with a starter agent, pairs it with your MindRoom Chat account, and starts it.
Or use the macOS app
brew install --cask mindroom-ai/tap/mindroomA native app that runs your agents on your Mac in the background, with one-click local models. Needs an Apple silicon Mac with macOS 14 or later.
Chat with your agents on the web, on the Mac, on iPhone and iPad, and on Android (in beta). Rather not run it yourself? Try hosted MindRoom.
These are stills from the showcase recordings: MindRoom Chat and a real MindRoom backend, with a fictional company and scripted model responses. More in the MindRoom README.
|
🔌 Connected to your tools and documents |
🔒 Private where it matters |
|
🧠 Memory that keeps improving |
🛡️ Safe to give real access |
|
💬 A chat app built for agents |
🌉 Works where you already are |
| Goal | Start here |
|---|---|
| Run MindRoom on your computer and chat in MindRoom Chat | Quick start |
| Run your agents on a Mac with the native app | macOS app guide |
| Self-host everything with Docker Compose | mindroom-stack |
| Give an agent its own NixOS machine (advanced) | lxc-nixos |
| Deploy for an organization on Kubernetes | Kubernetes guide |
| Let us run it for you | Hosted MindRoom |
| Learn configuration, tools, and deployment | docs.mindroom.chat |
| Project | What it does |
|---|---|
| mindroom | The agent runtime: orchestration, dashboard, 100+ tools, memory, knowledge bases, and the Matrix integration. |
| mindroom-chat | MindRoom's Cinny-based chat app built for agents, with threads, streaming replies, tool traces, approvals, canvases, and the agent's browser. |
| mindroom-stack | A Docker Compose stack with MindRoom, a Matrix homeserver, and MindRoom Chat. |
| lxc-nixos | A NixOS flake that runs the whole stack in an Incus container an agent can manage itself. |
| example-agents | An example agent workspace with a persona, operating rules, knowledge, memory, and skills. |
| matrix-mcp | A Matrix MCP server for local coding agents. |
| mindroom-egress-proxy | A network firewall and approval proxy for agent internet access. |
| agent-vault | MindRoom's fork of Infisical's Agent Vault, a credential broker that keeps API keys out of agent processes. |
| mindroom-nio | MindRoom's fork of matrix-nio, the Python Matrix client library the runtime is built on. |
| homebrew-tap | The Homebrew tap for the macOS app. |
Matrix and client forks
| Project | What it does |
|---|---|
| mindroom-tuwunel | MindRoom's fork of the Tuwunel Matrix homeserver, used by mindroom-stack. |
| mindroom-synapse | MindRoom's fork of Synapse for high-frequency edit streaming. |
| lk-jwt-service | MindRoom's fork of the MatrixRTC authorization service for voice calls. |
| mindroom-librechat | A LibreChat fork that shows MindRoom's server-side tool calls as native tool cards. |
Plugins add tools, skills, event hooks, OAuth providers, prompt enrichment, and safety policies without touching the core runtime, and they reload while MindRoom keeps running.
Install one with mindroom plugins install <name>, and see Plugins to configure and build them.
| Plugin | What it adds |
|---|---|
| agent-vault-bridge-plugin | Routes agent shell API calls through Agent Vault so credentials never enter the agent process. |
| deep-research-plugin | Runs bounded, multi-round web research on the agent's configured model and returns cited reports. |
| location-enrich-plugin | Adds real-time Dawarich location and movement context to agent prompts. |
| openviking-plugin | Provides automatic long-term memory extraction, recall, and management through OpenViking. |
| ping-hook-plugin | Shows the smallest useful event-hook plugin with a direct !ping-hook response. |
| response-audit-jev-plugin | Uses JEV or an LLM to check completed answers for citation and source-use issues, then requests corrections in the same thread. |
| restart-resume-plugin | Wakes tagged idle threads after MindRoom restarts, then clears their restart tags. |
| shell-guard-plugin | Intercepts and blocks configured dangerous shell tool calls before execution. |
| thread-goal-plugin | Keeps persistent thread goals available across context compaction and restarts. |
| thread-snooze-plugin | Resolves threads temporarily and wakes them automatically at a chosen time. |
| voice-enrich-plugin | Adds private speech-to-text guidance to prompts without changing visible messages. |
MindRoom is open source and built on Matrix, an open protocol with federation, end-to-end encryption, and a mature bridge ecosystem. Open an issue, read the contributing notes, or star the main repository to follow along.