What problem does it solve? Skills often fail because they trigger at the wrong time, overload context with duplicated reference material, or are never verified against real agent behavior. This Skill provides an authoring contract, quality gate, and testing methodology so skills trigger reliably, stay concise, and are validated against baselines before adoption. ## Core Features & Use Cases - Authoring contract and quality gate: Enforces discovery metadata rules (lowercase kebab-case name, 'Use when' descriptions), progressive disclosure structure, and link/ownership validation before a skill is considered ready. - TDD-style skill testing: Applies RED-GREEN-REFACTOR to skill development using pressure scenarios, verbatim rationalization capture, and loophole closing until agents comply under maximum pressure. - Reference library: Includes Anthropic best-practices digest, persuasion-principles research for compliance-critical skills, and a subagent pressure-testing guide. - Use Case: When a 900-line SKILL.md repeats three reference guides, use this Skill to move heavy material into references/, tighten the trigger description, and add eval prompts with baseline-vs-candidate evidence. ## Quick Start Use the writing-skills skill to review my new SKILL.md for trigger quality, context size, and missing evals.
