What problem does it solve? Writing effective agent skills is hard to get right on the first try, and without structured evaluation there is no way to know whether a skill actually improves agent behavior. This Skill provides a complete create-test-review-iterate loop with quantitative benchmarking and description optimization. ## Core Features & Use Cases - Guided Skill Authoring: Interviews the user to capture intent, then drafts a SKILL.md with proper frontmatter, progressive disclosure structure, and writing patterns. - Parallel Eval Runs with Baselines: Spawns with-skill and without-skill subagent runs, grades outputs against assertions, and aggregates pass rates, timing, and token usage into benchmark reports. - Interactive Review Viewer: Generates a browser-based viewer for qualitative feedback on outputs plus a quantitative benchmark tab, feeding user comments back into the next iteration. - Description Optimization: Runs an automated train/test loop that rewrites the skill's description to improve triggering accuracy on realistic queries. - Use Case: A user wants a skill that formats weekly sales reports. This Skill drafts it, runs three realistic test prompts with and without the skill, shows the results side by side, and iterates until the outputs pass review. ## Quick Start Help me create a new skill that converts meeting transcripts into structured action-item lists, then test and refine it.
