writing-skills

What problem does it solve? Writing process documentation for AI agents often fails because untested skills get rationalized away under pressure. This Skill applies Test-Driven Development to skill authoring: you run baseline scenarios without the skill, watch agents fail, write the skill to address those exact failures, and iterate until agents comply. ## Core Features & Use Cases - RED-GREEN-REFACTOR for documentation: Run pressure scenarios with subagents before writing any skill, capture verbatim rationalizations, and close loopholes iteratively. - Skill Discovery Optimization (SDO): Guidance on writing descriptions with triggering conditions only, keyword coverage, token efficiency, and naming conventions so agents actually find your skill. - Bulletproofing toolkit: Rationalization tables, red flags lists, persuasion principles, and form-matching rules to make discipline-enforcing skills resist workarounds. - Use Case: You want to create a new skill enforcing verification-before-completion. You first run a pressure scenario (time pressure + sunk cost + authority) without the skill, document how the agent skips verification, then write the skill targeting those exact excuses and re-test until compliant. ## Quick Start Ask the agent to help you create a new skill by first running a baseline pressure scenario with a subagent before writing any documentation.