MindRoom：在聊天中使用个人与团队AI代理

AI agents that know you and your work, in a chat app anyone can use. Open source, any model, self-host or hosted.

MindRoom 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 · Quick start 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 run Needs 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/mindroom A 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. See Quick start for every way to run it. See it in action Approve before it acts Actions you choose wait for your OK, with the exact arguments in view. The whole team, one thread Colleagues share an agent in a thread and see every answer stream in live. Canvases you can click An agent lays out free slots or a trip budget beside the chat and acts on what you pick. Data you can explore Leave a point out of a fit, then send the new result back to the agent. It drives, you take the wheel Watch an agent work in its own browser and take over when it needs your passkey. Connect once, then just ask Connect an account once, and the agent answers from it from then on. From your phone One voice note locks up the house; another becomes a reminder. Your morning, already handled A scheduled brief with today's meetings and what needs you. These are stills from the showcase recordings: MindRoom Chat and a real MindRoom backend, with a fictional company and scripted model responses. What people use it for Personal Plan a family trip, from flights to a packing list. Keep your calendar, reminders, and to-do lists in order, by voice. Keep notes, a journal, and memories you can find months later. Follow the topics you care about, with a digest only when something is new. Look after your homelab and smart home, with approval for the actions you choose. Build quick tools and scripts in the agent's own workspace. Work Find anything across email, chat, documents, tickets, and code, with sources. Get a morning briefing, or a summary of your week before a one-on-one. Write status updates from what actually happened in chat and the tracker. Turn a question into a report or a presentation built from your own documents. Triage the inbox and draft emails, approving each one before it is sent. Join a new team and ask its agent how the work fits together. Why MindRoom 🔌 Connected to your tools and documents Personal agents and shared team agents connect to 100+ tools, including email, calendar, Slack, Jira, GitHub, and any MCP server, and search your own documents. Even from a server across the world, they can use a computer you pair. 🔒 Private where it matters Pick a model per agent: a local one for your most personal data, a frontier one for coding. With local memory and your own server, nothing that agent sees leaves your home, and what you tell a private agent stays out of shared ones. 🧠 Memory that keeps improving Agents keep what matters from every conversation, get better as more people use them, and can turn work they repeat into reusable skills. 🛡️ Safe to give real access One-tap approval for the actions you choose, sandboxed code execution (on by default in hosted MindRoom), and opt-in end-to-end encryption on Matrix, the open standard governments use for secure messaging. 💬 A chat app built for agents MindRoom builds its own client for the web, Mac, iPhone, iPad, and Android (in beta), so agents can show live tool traces, ask for approval, open interactive canvases, join voice calls, and work in a real browser you can take over. 🌉 Works where you already are Bridges bring the same agents to Slack, Telegram, WhatsApp, and Discord, and an MCP gateway brings them to Claude Code and Codex. Full feature list Agents and teams Multi-agent orchestration — define specialist agents and teams in config.yaml or the dashboard; a built-in router picks the responder when you don't @-mention one, mentioning several agents makes them collaborate in a thread, and agents can hand tasks to each other or run reusable workflows (Teams, Agent Orchestration). Adaptive participation — in a thread with several people, agents you opt in that already replied there decide for themselves when an untagged message is theirs to answer, and a judge can decide whether a new message should interrupt the current task or wait (Threads). Personal rooms and access control — each person can get a private room with their own agent, and each agent answers only the people and rooms you allow (Personal Rooms, Access Control). Memory and learning Persistent memory — agents remember people, preferences, and context across conversations and platforms (Mem0 + ChromaDB, stored on your disk). Automatic skill learning — opt-in background reviews turn work an agent repeats into a small library of reusable skills, following the self-improvement loop of Hermes Agent (Skills). Automatic knowledge graph — an agent you opt in gets a Graphiti graph of the people, projects, and documents its conversations and tool results reveal, with when each fact held and where it came from, and a scheduled exploration can have it read the sources its tools and MCP servers reach to