What problem does it solve? It bridges the gap between a domain model (context-map with processes and bounded contexts) and executable work by generating a structured DAG of lean backend, frontend, sync, and infra tasks, so teams know exactly what to build and in what dependency order. ## Core Features & Use Cases - Process fan-out: Each process in the context-map becomes one BE produces task plus N FE consumes tasks, with the pairing declaring the operationId that a later contract step reconciles into OpenAPI. - dag.yaml generation: Emits a structure-only YAML file with release slices (vertical value increments) and depends_on/contract edges, while task state stays in the filesystem. - Lean task files: Produces 80-120 line task files with Gherkin acceptance criteria, invariant ACs from the tactical model, and anchored references instead of embedded state. - Use Case: After modeling a feature, run this to turn the context-map into slices like "recording viewable per machine" with BE/FE tasks wired by contract arcs, ready for a readiness gate. ## Quick Start Ask the agent to transform the context-map and architecture notes of the current feature into a dag.yaml with per-side lean tasks under the feature output directory.
