What problem does it solve? Bridging the gap between high-level planning and implementation is hard: epics and sprints exist, but teams lack a structured domain model, documented architecture decisions, and implementation-ready task specifications. This Skill guides the AI-DLC Construction Phase by producing Domain Designs, Logical Designs, Architecture Decision Records (ADRs), and validated Task Specifications for approved epics. ## Core Features & Use Cases - Domain and Logical Design: Applies DDD principles (bounded contexts, aggregates, entities, domain events) and layered technical guidance (global, stack-specific, project-level) to propose architecture patterns with deviation analysis against existing codebase patterns. - ADR Generation: Creates Architecture Decision Records for significant decisions, including deviations from organizational standards, with context, consequences, and alternatives. - Task Specification with Quality Gates: Generates implementation-ready task specs with concrete acceptance criteria, data contracts, and error handling, enforced by confidence scoring, detail-sufficiency gates, and publish approval checkpoints. - Multi-Backend Storage: Publishes artifacts to GitLab (markdown files), Linear (design doc issues), or Confluence (child pages) based on automatic backend detection. - Use Case: After running epic decomposition for a new authentication feature, invoke this Skill to model the domain, select architectural patterns, record ADRs, and produce sprint-grouped task specifications ready for team review. ## Quick Start Ask the AI to create the domain model and logical design for your approved epic, for example by saying: create the domain design and ADRs for the authentication epic.
