A persistent tmux-based terminal server for AI agents, enabling real-time spectating and SSH host-group dispatch. It provides a structured Model Context Protocol (MCP) server experience with centralized session management and remote monitoring capabilities.
🛠️ Key Features
- Persistent tmux terminal sessions for AI agents
- Real-time spectating of running sessions
- SSH host-group dispatch for targeted access
- Lightweight MCP server oriented around model context
🚀 Use Cases
- Observing AI agent workflows in live sessions
- Dispatching commands to specific SSH host groups
- Centralized context sharing for model-driven experiments
⚡ Developer Benefits
- Clear MCP-based server interface for integration
- Docker-friendly and easy to deploy in CI/CD
- Spectator mode enables safe, read-only monitoring
⚠️ Limitations
- Details on configuration and scalability are limited in the provided data
- No explicit performance or security notes in the source excerpt