What problem does it solve?
Coordinating AI coding agents across a multi-task development pipeline is chaotic: tasks get lost, progress goes unreported, dependencies break, and parallel agents work without observability. This Skill gives developer agents a structured workflow to claim tasks, execute work, self-check acceptance criteria, and submit for human verification within the Chorus AI-DLC platform.
Core Features & Use Cases
- Full Task Lifecycle Management: Claim, start, report, self-check acceptance criteria, and submit tasks for admin verification using Chorus MCP tools.
- Parallel Subagent Orchestration: Spawn up to 4 concurrent Kiro subagents with session-based observability, wave-based DAG dependency resolution, and per-session attribution.
- Review Gateway Integration: Trigger independent task-reviewer and code-reviewer subagents that post VERDICT comments before work ships.
- Use Case: A PM agent creates a proposal with 8 tasks in a dependency DAG. The main agent spawns subagents wave by wave, each claiming tasks, reporting commits and PRs, self-checking acceptance criteria, and submitting for verification until the idea ships.
Quick Start
Ask the agent to check in with Chorus, list available tasks for the project, claim the highest-priority unblocked task, and start development following the chorus-develop workflow.