Mattermost MCP Client

What is Mattermost MCP Client?

A Python-based MCP (Model Context Protocol) client that connects Mattermost with MCP servers, enabling command execution and tool management through Mattermost channels. It runs on Python 3.13.1+ and requires a Mattermost server with a bot account.

How to use Mattermost MCP Client?

Install by creating a virtual environment with uv venv, then uv sync. Configure via .env file with Mattermost URL, token, team, channel, and MCP server settings. Start the client with python src/mattermost_mcp_client/main.py. Interact using Mattermost commands like !mcp help, !mcp servers, !mcp <server_name> call <tool_name> <args>. Optionally use the mcp_tool_caller.py utility for command-line testing.

Key features of Mattermost MCP Client

• Connects Mattermost with MCP servers via stdio.
• Supports multiple MCP servers (simple, Ollama, Mattermost API).
• Execute MCP tools directly from Mattermost channels.
• List servers, tools, resources, and prompts.
• Command-line utility for testing MCP servers.
• Customizable command prefix (default !mcp).

Use cases of Mattermost MCP Client

• Run AI model queries (e.g., generate text via Ollama) through Mattermost.
• Post messages to Mattermost channels programmatically.
• Automate tasks by combining MCP tools from multiple servers.
• Test MCP server tools without leaving the command line.

FAQ from Mattermost MCP Client

What does Mattermost MCP Client do?

It acts as a bridge between Mattermost and MCP servers, allowing users to call MCP tools and manage servers via Mattermost channel commands.

What platforms or models does it support?

It runs on Python 3.13.1+. It supports MCP servers configured via stdio, including Ollama for local LLM models and custom Python-based servers.

How do I configure MCP servers?

Edit the mcp-servers.json file in src/mattermost_mcp_client/ to define server commands, arguments, and type (stdio). Environment variables in .env set Mattermost connection details.

Does it support npx-based MCP servers?

Not yet; support for npx-based MCP servers is listed as a next step.

Are there any known limitations?

Currently missing authentication for secure communication, file uploads, and tool-calling agent orchestration. These are planned future features.