Control

•
3
•

Execute commands, orchestrate workflows, and transfer files across distributed Windows, Linux, and macOS machines via peer-to-peer connections and MCP integration.

Control establishes a secure peer execution network across Windows, Linux, and macOS systems, connecting private fleets through a central gateway while communicating directly over WebRTC with mutual TLS. It empowers developers and AI agents to coordinate durable tasks, stream logs, manage TCP tunnels, and exchange SHA-256 verified artifacts across hybrid user and system contexts without relying on public cloud bottlenecks. Featuring native Model Context Protocol (MCP) support, terminal dashboards, and fine-grained authority delegation, Control serves as a robust foundation for distributed computing and agent-driven automation.

主な機能

01Direct WebRTC peer-to-peer connections with WebSocket relay fallback and mutually authenticated TLS
02Durable remote task execution with idempotent submission, cancellation, and live log streaming
03Resumable peer-to-peer artifact delivery with automatic SHA-256 verification
04Native Model Context Protocol (MCP) integration and AI skill support for popular AI clients
05Cross-platform execution across Windows, Linux, and macOS with isolated user and system profiles
063 GitHub stars

ユースケース

01Empowering AI agents to inspect, control, and execute commands across a fleet of physical or virtual machines
02Distributing compute-intensive workflows and exchanging large artifacts directly between workstations and build servers
03Creating secure forward or reverse TCP tunnels and automating desktop actions across remote environments