What problem does it solve?
Team-based AI work becomes slow and error-prone when you try to coordinate many specialized steps (planning, implementation, verification, and fixes) manually instead of letting agents manage dependencies and handoffs.
Core Features & Use Cases
- Parallel teammate execution with dependencies: Decomposes a task into subtasks, assigns owners, tracks blocked/unblocked work, and coordinates progress across N agents.
- Staged pipeline orchestration: Runs a canonical flow (team-plan → team-prd → team-exec → team-verify → team-fix loop) with stage-appropriate specialist routing.
- Resume/cancel aware state management: Persists team phase and history so work can resume after interruption, and supports controlled cancellation with orderly shutdown.
- Optional Ralph persistence loop integration: Wraps the team pipeline with retry/persistence behavior when you request the ralph modifier.
Quick Start
Run /oh-my-claudecode:team 5:executor "fix all TypeScript errors across the project" to have coordinated agents plan, execute, verify, and iterate until the fix criteria are met.