What problem does it solve? Writing an effective skill is hard to get right on the first try: descriptions undertrigger, instructions overfit to a few examples, and there is no systematic way to know whether a change actually helped. This Skill provides a structured draft-test-review-improve loop with quantitative benchmarks so skill quality is measured rather than guessed. ## Core Features & Use Cases - Guided Skill Authoring: Interviews the user about intent, trigger contexts, and output formats, then writes a SKILL.md following progressive-disclosure and writing-style best practices. - Parallel Eval Runs with Baselines: Spawns with-skill and without-skill (or old-version) subagent runs for each test prompt, captures timing and token data, and grades outputs against assertions. - Benchmark Aggregation and Review Viewer: Aggregates grading results into benchmark.json/benchmark.md with mean, stddev, and deltas, then launches an HTML viewer for qualitative feedback. - Description Optimization Loop: Generates realistic should-trigger/should-not-trigger queries, reviews them in an HTML template, and runs an automated train/test optimization loop to improve triggering accuracy. - Use Case: You drafted a skill that converts meeting notes into Jira tickets but it only triggers half the time. Use this Skill to run trigger evals, optimize the description, and verify the improvement on a held-out test set. ## Quick Start Help me create a new skill that turns my weekly status notes into a formatted report, then test it with a few realistic prompts.
