What problem does it solve? Translating a Product Requirements Document into an actionable technical design is error-prone when done ad hoc: architecture assumptions go unexamined, decisions are not recorded, and the resulting plan drifts from the existing codebase. This Skill enforces a structured, interactive process that produces a codebase-informed Technical Specification with documented Architecture Decision Records. ## Core Features & Use Cases - Context-Grounded Design: Reads the existing PRD and ADRs, then explores the codebase architecture before proposing any technical approach. - Interactive Clarification: Asks targeted one-at-a-time questions on architecture, data models, APIs, and testing, requiring user approval before writing the final document. - ADR Generation: Records every significant technical decision as a numbered ADR file with alternatives considered and trade-offs. - Use Case: After finishing a PRD for a new scheduling feature, run this Skill to produce a _techspec.md with system architecture, core interfaces, API endpoints, impact analysis, and a dependency-ordered build sequence ready for task breakdown. ## Quick Start Ask the AI to create a technical specification for the feature named 'scheduling' using the existing PRD in the task directory.
