Background Job

Enables coding agents to execute long-running shell commands asynchronously with full process management capabilities.

MCP server implementation by Dylan Navajas Gluck that enables asynchronous execution and management of long-running shell commands with full process lifecycle control. Built with Python 3.12+ and FastMCP, the implementation provides seven core tools for background job management: execute commands and get job IDs, monitor status and output with real-time tailing, send input to interactive processes via stdin, and terminate running jobs with graceful cleanup. Features ring-buffered output capture with configurable size limits, concurrent job management with resource constraints, security validation against dangerous command patterns, and comprehensive process monitoring including PID tracking and exit code handling. Designed for development workflows involving build processes, test suites, development servers, or any scenario where AI assistants need to start long-running operations and monitor their progress without blocking on command completion.