What problem does it solve? Raw intent — vague ideas, brain dumps, PRDs, meeting transcripts — is too unstructured for downstream planning and implementation skills to consume reliably. This Skill condenses any input into a canonical SPEC.md kernel (Why, Capabilities, Constraints, Non-goals, Success signal) plus companion files, so every downstream BMad skill works from one machine-readable contract. ## Core Features & Use Cases - Spec distillation: Converts any intent input into the five-field kernel, routing overflow content (glossaries, diagrams, entity catalogs) into named companion files. - Append-only memory: Records every decision, capability, and assumption in .memlog.md and re-derives SPEC.md from it, so updates never drift or lose stable capability IDs. - Validation and story breakdown: Self-validates against eight Spec Law rules, and can interactively break a spec into a stories.yaml execution plan. - Use Case: Paste a messy customer email thread and receive a spec folder with SPEC.md, a glossary companion, and a memlog recording every decision and open question. ## Quick Start Ask the assistant to distill the attached PRD or idea notes into a BMad spec with a slug like "quarter-drop".