rocketchat-mcp
MCP server for Rocket.Chat, enabling AI agents to interact with Rocket.Chat workspaces via tools like listing users, sending messages, and managing channels.
README
MCP Rocket.Chat Setup Guide
Rocket.Chat MCP server for Nanobot, Cursor, and Claude Desktop. Use it as:
- HTTP server (Docker) — no credentials in the container; clients (e.g. Nanobot) send credentials in request headers.
- Stdio — credentials from environment or CLI args.
Credentials
When using HTTP (recommended with Nanobot in Docker): the MCP server runs without credentials. Pass them in Nanobot’s config under mcpServers.<name>.headers; they are sent with every request. No .env or container env needed.
When using stdio or optional Docker defaults: use environment variables or CLI args.
Use either username/password or user ID + auth token (token takes precedence when both are set).
| Source | Server URL | Username / Password | User ID / Auth Token |
|---|---|---|---|
| Headers | X-RocketChat-Server-URL or X-RocketChat-URL |
X-RocketChat-Username, X-RocketChat-Password |
X-RocketChat-User-Id, X-RocketChat-Auth-Token |
| Env | ROCKETCHAT_SERVER_URL or ROCKETCHAT_URL |
ROCKETCHAT_USERNAME, ROCKETCHAT_PASSWORD |
ROCKETCHAT_USER_ID, ROCKETCHAT_AUTH_TOKEN |
| CLI | --server-url |
--username, --password |
(not available) |
Run as MCP inside Nanobot via HTTP (recommended)
The MCP server runs without credentials in the container. Credentials are sent by Nanobot on each request via headers.
-
Start the MCP server (no env vars or
.envrequired):docker compose up -dIt listens at
http://<host>:3011/mcp. -
Configure Nanobot (
~/.nanobot/config.json): add the server with url and headers (credentials live only in your config):{ "tools": { "mcpServers": { "mcp-rocketchat": { "url": "http://rocketchat-mcp:3011/mcp", "headers": { "X-RocketChat-Server-URL": "https://chat.example.com", "X-RocketChat-User-Id": "your-user-id", "X-RocketChat-Auth-Token": "your-auth-token" } } } } }Alternative — username/password (server will log in on first use):
"headers": { "X-RocketChat-Server-URL": "https://chat.example.com", "X-RocketChat-Username": "your-bot-user", "X-RocketChat-Password": "your-password" }Use the hostname that reaches the MCP server from Nanobot (e.g.
rocketchat-mcpif both are on the same Docker network, orhost.docker.internal:3011if the server runs on the host). -
Restart Nanobot. Credentials are sent with every MCP request; nothing is stored in the container.
Run as MCP inside Nanobot (stdio)
Nanobot starts the server as a subprocess and talks over stdio. Credentials must be available as env vars where Nanobot runs.
-
Set environment variables (e.g. in your shell profile or where you start Nanobot). Use either username/password or user ID + auth token:
export ROCKETCHAT_SERVER_URL="https://your-server.com" # Option A: username/password export ROCKETCHAT_USERNAME="your-bot-user" export ROCKETCHAT_PASSWORD="your-password" # Option B: user ID + auth token (instead of A) # export ROCKETCHAT_USER_ID="your-user-id" # export ROCKETCHAT_AUTH_TOKEN="your-auth-token" -
Add the MCP server to Nanobot config (
~/.nanobot/config.json):{ "tools": { "mcpServers": { "rocketchat": { "command": "python", "args": ["/absolute/path/to/rocketchat.py"] } } } }Or with
uvfrom the project directory:{ "tools": { "mcpServers": { "rocketchat": { "command": "uv", "args": ["run", "--project", "/path/to/rocketchat-mcp", "rocketchat.py"] } } } } -
Restart Nanobot (e.g.
nanobot gateway). The RocketChat tools will be available to the agent.
No --transport or --server-url is needed: the server uses stdio by default and reads credentials from the environment.
Run with Cursor / Claude Desktop (stdio)
Same as Nanobot: use env vars for credentials and start the server with no --transport (stdio is default).
Example (credentials in args; avoid in shared environments):
{
"mcpServers": {
"rocketchat": {
"command": "python",
"args": [
"C:/path/to/rocketchat.py",
"--server-url", "https://your-server.com",
"--username", "user",
"--password", "pass"
]
}
}
}
Prefer setting credentials in the environment (ROCKETCHAT_SERVER_URL plus either username/password or ROCKETCHAT_USER_ID/ROCKETCHAT_AUTH_TOKEN) and using only:
{
"mcpServers": {
"rocketchat": {
"command": "python",
"args": ["C:/path/to/rocketchat.py"]
}
}
}
Docker: no credentials in the container
The image is built and run without any RocketChat credentials. Start with:
docker compose up -d
The server listens on http://0.0.0.0:3011/mcp. Clients (e.g. Nanobot) must send credentials on each request via the headers listed in the credentials table. Optionally you can set ROCKETCHAT_* env vars in the container if you want a default for clients that don’t send headers.
Local Python setup (no Docker)
uv venv
.venv\Scripts\activate # or: source .venv/bin/activate
uv sync
-
Stdio (for Nanobot/Cursor): set
ROCKETCHAT_*env vars, then run:python rocketchat.py(Nanobot/Cursor will start this for you when configured as above.)
-
HTTP (local test):
set MCP_TRANSPORT=streamable-http python rocketchat.pyOr:
python rocketchat.py --transport streamable-http --port 3011
Tools exposed to the agent
list_users– list usersfind_channel– find room(s) by name (channels, DMs, private); returns room_id for messages/sendlist_all_rooms– list all rooms (channels, DMs, private, etc.)send_message_in_channel– send a message (channel name or room_id)get_channel_messages– get recent messages in a room (requires room_id)get_user_info– info for a usercreate_channel– create a channel
To get the last message in channel XY: call find_channel("XY") first, then get_channel_messages(room_id=<from find_channel>, count=1).
For more details, see Rocket.Chat API and Nanobot README.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.