What problem does it solve? Complex work that spans three or more interdependent units often fails from unclear sequencing, ambiguous authority, or unverified intermediate artifacts. This Skill structures such work as a bounded task graph with explicit dependencies, acceptance checks, budgets, and stop conditions so each node is verifiable before dependents unblock. ## Core Features & Use Cases - Bounded task graphs: Define each node's instruction, authority, dependencies, expected artifact, acceptance check, budget, and failure policy, then route work units through the co-packaged coordinator runtime. - Independent review routing: Bind caller-verified independent reviewers to governance or review nodes and verify returned lineage before accepting independent evidence. - Vertical slice decomposition: Split product-feature implementation into tracer-bullet slices that each cut through all layers, with expand-contract sequencing reserved for wide refactors. - Use Case: A user asks to ship a feature touching schema, API, and UI. The Skill decomposes it into demoable vertical slices with explicit dependencies, dispatches each to a provider via one bounded routing request, and gates integration on verified artifacts. ## Quick Start Ask the agent to orchestrate this feature as a task graph with explicit dependencies and acceptance checks for each node.