Design clear, versioned, backward-compatible interfaces and contracts between units, services, and external consumers. Applies wherever interfaces/contracts are designed or reviewed.
Evaluate software architecture for sound boundaries, intentional coupling, explicit trade-offs, scalability, deployability, operability, and failure behaviour. Applies wherever architectural decisions are designed or reviewed.
Identify and define the core domain concepts of a system — components, entities, relationships, ownership boundaries, and the language the system speaks. Applies wherever domain structure, entity ownership, or business capability boundaries are designed or reviewed.
Apply disciplined implementation practices across backend, frontend, infrastructure, configuration, and tests: write-test-verify, repository convention matching, integration boundary control, and appropriate test strategy.
AI-DLC workspace kickoff. Handles the welcome banner and workspace setup — creating the intent directory, state file, and audit file. Read by the orchestrator at the start of every new intent.
AI-DLC workflow orchestrator. Activate whenever the user states a fresh development intent — building, creating, implementing, fixing, migrating, refactoring, or adding a feature to a codebase. Use this skill for any free-form development prompt such as "build X", "add feature Y", "migrate Z to W", "fix the bug in V", "refactor U", or "create a new service for T".
Elicit, decompose, and structure requirements from ambiguous human intent into verifiable, traceable specifications. Applies wherever requirements are discovered, refined, validated, or used to check downstream artifacts.
Evaluate product artifacts — requirements, user stories, personas, and wireframes — for clarity, value, testability, traceability, and scope discipline. Applies wherever product-facing work is reviewed before it moves downstream.