SSH Tmux is a Model Context Protocol (MCP) server that manages persistent SSH sessions for AI agents using a local `tmux` instance. It provides visual screen snapshots for natural AI interaction and state tracking.

Do SSH sessions persist if the server or AI client restarts?

Yes, a core feature of SSH Tmux is that SSH connections remain alive within `tmux` even if the MCP server or your AI client restarts, ensuring continuity for long-running tasks.

What are the main benefits for AI agents?

It offers persistent SSH connections, real-time observability for AI activity, robust command execution, and native file transfer capabilities, simplifying complex automation and interaction with remote systems.

Can I monitor what an AI agent is doing in real-time?

Yes, you can manually attach to the underlying `tmux` session (e.g., `tmux attach -t mcp-ssh`) to observe the AI agent's actions, commands, and output in real-time, just like a human.

SSH Tmux

Name: SSH Tmux
Author: devnullvoid

bydevnullvoid

•

Developer Tools

Productivity & Workflow

Deployment & DevOps

Manages persistent SSH sessions for AI agents using a local `tmux` instance, providing visual screen snapshots for natural interaction.

This high-performance Model Context Protocol (MCP) server addresses the limitations of traditional SSH automation for AI agents. Instead of relying on complex regex patterns or stateless command execution, mcp-ssh-tmux leverages a local `tmux` instance to maintain persistent SSH connections. This innovative approach allows AI agents to interact with remote environments by interpreting visual screen snapshots, similar to human interaction, thereby eliminating parsing complexity and enabling more reliable, stateful operations on remote systems.

Key Features

01File Transfer: Enables native reading and writing of remote files using `cat` and `tee` over the existing PTY.

02Persistence: SSH connections remain alive in `tmux` even if the MCP server or AI client restarts.

03Safety: Includes built-in command validation to prevent common dangerous operations.

04Observability: Manually attach to `tmux` sessions to monitor AI agent activity in real-time.

05Reliability: Uses `ssh -G` for robust configuration resolution, handling aliases and identity files.

061 GitHub stars

Use Cases

01Enabling AI agents to develop, debug, and deploy applications on remote environments.

02Automating complex, stateful interactions with remote servers for AI agents.

03Providing a reliable and observable platform for AI agents to manage remote infrastructure.