Orchestrates large features as RFC-driven multi-agent DAG work units with quality gates and a merge queue.
- What it does
- This skill defines a pipeline for breaking a large feature request into independently verifiable work units arranged as a dependency graph. It specifies unit fields such as id, depends_on, scope, acceptance tests, risk level and rollback plan, plus complexity tiers and a per-unit quality pipeline from research through merge-readiness reporting. It also covers merge queue rules and recovery steps for stalled units. Outputs include an RFC execution log, unit scorecards, dependency graph snapshots and an integration risk summary.
- When to use it
- Use it when a feature is too large for a single agent pass and must be split into separate, independently verifiable work units. It suits multi-unit orchestration where dependencies, quality gates and integration risk need explicit management.
- Requirements
- Instructions only; no scripts are shipped. It assumes an agent environment capable of running multiple work units, branches and integration tests, but names no specific tools, packages or credentials.