What problem does it solve? Authoring a new agent skill for a published analytic method requires many mandatory sections — frontmatter, invocation rules, procedure, output template, worked example, verification checklist, and eval cases — and missing any of them breaks consistency across the library. This template gives authors a complete, annotated skeleton so every new skill meets the same structural bar. ## Core Features & Use Cases - Complete SKILL.md skeleton: Includes YAML frontmatter, when-to-invoke routing with sibling near-misses, a numbered procedure, an output template, a worked example, a verification checklist, anti-patterns, and references. - Eval scaffolding: Ships an evals/evals.json template with positive and negative (near-miss) test cases for discoverability testing. - Use Case: A contributor adding a new structured analytic technique (e.g., a decision-analysis method) copies this directory, replaces the placeholders with the method's published labels and steps, and ships a spec-compliant skill with evals. ## Quick Start Copy the skills/_TEMPLATE directory to a new skill folder and fill in every placeholder with the published method's name, steps, output template, and eval cases.
