What problem does it solve? When a product area contains user stories spanning multiple surfaces (UI, API, MCP, CLI), you would otherwise have to figure out which per-surface test-generation skill applies to which stories and invoke each one manually. This dispatcher walks every story in the area, groups them by effective surface, and routes each group to the right sibling skill in one pass. ## Core Features & Use Cases - Surface-aware grouping: Determines each story's effective surface from its explicit surface field or by inferring it from the story-key prefix (API-, MCP-, CLI-), with an unspecified bucket for anything else. - Routing plan with confirmation: Presents a routing table showing which sibling skill (defprod-create-ui-tests, defprod-create-api-tests, defprod-create-mcp-tests, defprod-create-cli-tests) will handle which stories, and asks before invoking. - Resilient orchestration: Invokes siblings in a stable ui → api → mcp → cli order, continues past individual sibling halts, and produces a final summary including deferred stories and inference-routed candidates for surface backfill. - Use Case: You have a STATUS area where STATUS-01 is an API story, STATUS-02 is UI, and STATUS-03 is CLI. Run the dispatcher once and it generates Playwright, Vitest, and Jest subprocess tests via the matching sibling skills without you picking them yourself. ## Quick Start Ask your agent to run /defprod-create-area-tests with your product name and area key, for example: generate tests for every story in the STATUS area of my product.
