# Ternimal MCP server

> Provides a secure remote terminal that allows AI agents to execute commands and manage files on remote hosts via the Model Context Protocol.

- Repository: https://github.com/ternimal/tern
- Page: https://hostedmcp.org/servers/ternimal/tern
- Transport: —
- Stars: 1
- License: AGPL-3.0

## Environment variables

| Name | Required | Secret | Description |
|---|---|---|---|
| `TERN_TOKEN` | no | no |  |

## FAQ

### How can AI agents use Ternimal?

AI clients like Claude Code or Cursor can work on your agents through the Model Context Protocol (MCP) using the `tern mcp` command.

### Is the connection secure?

Yes, client-to-agent traffic uses end-to-end TLS with mandatory authentication, ensuring relays only forward opaque ciphertext.

### Can I use Ternimal to reach a machine behind a NAT?

Yes, you can reach an agent behind NAT by using a relay to bridge the connection.

### What are the default permissions for AI agents?

By default, the MCP server is read-only and will ask for your approval before performing chosen actions.