Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Added attribution at the top of SKILL.md linking to the original mcp-ssh-manager repository and author. - Clarified that the skill provides documentation, workflows, and best practices for MCP ssh-manager usage. - No changes to functionality or commands; content and usage examples remain the same. - Version unchanged in SKILL.md and metadata.
What This Skill Does
Connects AI agents to remote servers over SSH using the mcp-ssh-manager MCP server. Supports command execution, file transfers, port tunneling, session management, and server monitoring through a consistent set of tools. Results can be persisted to a local workdir structure for comparison and audit.
Maintains stateful SSH sessions across multi-step workflows, avoiding the overhead of re-authenticating for each command.
When to use it
- Deploy a Node.js app to a remote server via persistent SSH session
- Tail nginx access logs and filter for errors in real time
- Create a local port forward to access a remote database
- Schedule nightly file backups on a remote host
- Run a health check across CPU, memory, and running services
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Deploy the latest build to the production server
- 1Call ssh_session_list to check for an existing session on the target server
- 2Start a new session with ssh_session_start server="prod" name="deploy"
- 3Send commands via ssh_session_send to pull the latest code, install dependencies, and build
- 4Upload the compiled dist folder using ssh_upload to the remote web root
- 5Restart the application process, then close the session with ssh_session_close
Application is updated and running on the remote server; session is cleanly closed
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.