Orchestrate

Automating AI agent task distribution requires analyzing GitHub Project tasks for readiness and dependencies, then creating wave-based execution plans. This Skill coordinates parallel agent execution by spawning isolated worktrees and generating self-contained prompts for autonomous task completion.

Agent isolation works by utilizing Git worktrees to separate concurrent AI agent environments during task execution. This Skill further manages agent isolation through optional Docker sandboxes, ensuring autonomous agent authentication and deterministic task execution without environmental conflicts.

You need both the GitHub CLI (gh) and Docker installed to fully utilize multi-agent orchestration. The GitHub CLI manages project board tasks and pull requests, while Docker provides the optional sandbox environments required for secure, isolated agent execution.

Monitoring AI agent progress and cleaning up stale worktrees is handled by tracking active task execution waves and identifying abandoned branches. This Skill facilitates the merging of completed pull requests and manages the teardown of Docker sandboxes and worktrees after task completion.

Wave-based planning supports parallel AI agent execution within a GitOps framework by structuring task distribution into sequential execution phases. This Skill automatically assesses task dependencies on your project board to determine readiness and creates structured execution waves for autonomous agents.

Autonomous agent authentication in task distribution environments relies on proper Git worktree setup and GitHub CLI configuration. This Skill ensures agents operate autonomously within their isolated environments by managing authentication contexts and communication channels throughout the task lifecycle.