
from agent-process-guard21
Ensures terminal process hygiene for coding agents, preventing orphaned background processes and blocking terminal execution.
Process Guard transforms terminal execution into a managed lifecycle. It classifies commands as finite, long-running, or interactive, ensuring that servers and watchers are run as managed background processes with PID tracking and verified readiness checks.
Activate this skill before running any terminal command that might remain alive (e.g., npm run dev, docker compose up, or database migrations) to prevent the agent from blocking or leaving zombie processes.
Primarily targeted at Codex and Claude Code, but compatible with any Agent Skills-compliant coding agent.
This skill has not been reviewed by our automated audit pipeline yet.