What problem does it solve? Creating a new AI skill by hand leads to inconsistent naming, vague discovery descriptions, and duplicated workflows. This Skill enforces a single authoring convention so every new skill is discoverable, correctly structured, and non-overlapping with existing ones. ## Core Features & Use Cases - Template-based scaffolding: Runs init_skill.py to copy the canonical SKILL.md template while enforcing naming rules (lowercase hyphenated, max 64 chars, no reserved words). - Mechanical validation: Runs validate_skill.py to check frontmatter, name-directory match, description triggers, body length, inline block size, and nested reference depth. - Overlap prevention: Greps existing skills for trigger phrases before writing anything, stopping near-duplicate skills at the source. - Use Case: When you ask to "make a skill for scaffolding Convex tables", it clarifies the brief, checks for overlap, scaffolds the directory, validates it, and reports the result in one sentence. ## Quick Start Ask the assistant to create a new skill for your repeated workflow, describing its purpose and trigger phrases.