What are the prerequisites for running A Better Mcp ssh Server?
Go 1.25.0 or later must be installed, along with SSH (tested on macOS). A `~/.ssh/config` file is required to define host entries.
@zhangyc310
Basic information
A Better Mcp ssh Server.
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"ssh": {
"command": "/your_path_to_mcpssh/mcpssh",
"args": []
}
}
}Initiates a new SSH connection to a specified host using a PTY. Returns a `session_id`.
Sends input to and reads output from an active SSH session. This allows for interactive shell usage (e.g., handling prompts, running scripts).
Terminates an active SSH session and cleans up resources.
A Better Mcp ssh Server is a Go-based MCP server that manages SSH sessions for AI agents. It enables an AI to interact with remote servers via SSH by creating persistent sessions, sending commands, and reading outputs through a pseudo-terminal (PTY) interface. It reads hosts from ~/.ssh/config, so users do not need to list every host in the MCP configuration file.
Build the binary with go build -o mcpssh (requires Go 1.25.0+ and SSH installed). Then configure your MCP client (e.g., Codex or Gemini) to run the binary as a command server without arguments. After setup, the AI can connect to any host defined in ~/.ssh/config by name using the provided tools.
~/.ssh/config automaticallyGo 1.25.0 or later must be installed, along with SSH (tested on macOS). A `~/.ssh/config` file is required to define host entries.
It reads host definitions from `~/.ssh/config`. There is no need to add every host to the MCP configuration file; the AI can reference hosts by their config alias.
Three tools: `start_session` (initiates SSH connection and returns a session ID), `interact_session` (sends input to and reads output from the PTY), and `close_session` (terminates the session and cleans up resources).
The server depends on `github.com/mark3labs/mcp-go` (MCP SDK), `github.com/creack/pty` (PTY management), and `github.com/google/uuid` (session ID generation).
Each `start_session` call creates a persistent PTY-based SSH connection and returns a unique session ID. The AI uses `interact_session` to send commands and read outputs, and `close_session` to terminate the session.

Routebase keeps your API's design, docs, contract tests, mocks and monitoring in one living OpenAPI spec so they can't drift apart. Agents work on that same source of truth, under your team's permissions.

Know when your AI agents, cron jobs and CI pipelines fail, stall or go quiet. Your agent sets up monitors from a plain description, tells you what is broken and why, reads run history and traces, and writes notes on inci
Modelglass turns "which model should I use?" into a lookup instead of a guess. Seven tools give your agent live pricing across image, language, video, and audio models. Sourced and versioned, not scraped once and forgott


Word search generator, crossword generator, and sudoku generator + solver as a local-first MCP server. 15 deterministic tools: printable PDF puzzle worksheets, themed word banks, and verifiable LLM evals. From the makers
Comments