Model Context Protocol server that lets LLMs collaboratively drive tmux
{
"mcpServers": {
"io-github-k8ika0s-mcp-tmux": {
"args": [
"-y",
"@k8ika0s/mcp-tmux"
],
"command": "npx"
}
}
}Are you the author?
Add this badge to your README to show your security score and help users find safe servers.
Model Context Protocol server that lets LLMs collaboratively drive tmux
Is it safe?
No known CVEs for @k8ika0s/mcp-tmux.
No authentication — any process on your machine can connect.
License not specified.
Is it maintained?
Last commit 27 days ago. 1 stars.
Will it work with my client?
Transport: stdio, sse, http. Works with Claude Desktop, Cursor, Claude Code, and most MCP clients.
Run this in your terminal to verify the server starts. Then let us know if it worked — your result helps other developers.
npx -y '@k8ika0s/mcp-tmux' 2>&1 | head -1 && echo "✓ Server started successfully"
After testing, let us know if it worked:
No known vulnerabilities.
This server is missing a description. Tools and install config are also missing.If you've used it, help the community.
Add informationHave you used this server?
Share your experience — it helps other developers decide.
Sign in to write a review.
Dynamic problem-solving through sequential thought chains
A Model Context Protocol server for searching and analyzing arXiv papers
An open-source AI agent that brings the power of Gemini directly into your terminal.
The official Python SDK for Model Context Protocol servers and clients
MCP Security Weekly
Get CVE alerts and security updates for io.github.k8ika0s/mcp-tmux and similar servers.
Start a conversation
Ask a question, share a tip, or report an issue.
Sign in to join the discussion.
Remote-first tmux co-pilot for humans + LLMs.
mcp-tmux is a Model Context Protocol server for people who learned the hard way that large language models are extremely confident about terminals they cannot actually see. It lets an LLM operate inside tmux with real, inspectable state instead of vibes, screenshots, or remembered lies. The server anchors the model to concrete reality: SSH-aware session discovery and bootstrapping, authoritative window and pane topology, and pull-based state snapshots so context is fetched when needed rather than hallucinated continuously. It can spawn or reattach to remote tmux sessions via your SSH config, inject keystrokes into the correct pane on purpose, read scrollback with full historical context, manage windows and splits deterministically, and enforce defaults so the model doesn’t calmly execute commands in a pane that stopped existing five minutes ago.
You need this if you want an LLM to assist in real terminals without pretending the terminal is a chat log. This is for pairing with an AI on live systems, debugging multi-pane chaos, automating the boring parts while you handle the judgment calls, or letting a model help without giving it omnipotent shell access and hoping nothing exciting happens. mcp-tmux exists to replace guesswork with state, narration with observation, and “I think you’re in the left pane” with provable fact.
Disclaimer: reasonable engineering effort has been applied to reduce the probability of reality-ending command execution. Guardrails exist. So do sharp edges. If something catastrophic happens, it will almost certainly be because a human approved it, ignored context, or decided this was a good idea at the time. Logs are kept. History remembers. Responsibility remains firmly biological.
npm install
npm run build
MCP_TMUX_HOST=my-ssh-alias MCP_TMUX_SESSION=collab npx @k8ika0s/mcp-tmux
tmux_open_session → tmux_default_context → tmux_list_windows / tmux_list_panes → tmux_send_keys → tmux_capture_pane
tmux_state.