What problem does it solve? Effective prompts discovered during work sessions are often lost, forcing you to reinvent prompt patterns that already worked. This Skill captures successful prompts, generalizes them into reusable templates, and appends them to a structured catalog file. ## Core Features & Use Cases - Session Scanning: Identifies prompts that achieved the desired outcome, used clever techniques, or demonstrated good prompt engineering. - Categorization & Templating: Classifies prompts into CODE, REVIEW, DOCS, TEST, DEPLOY, DESIGN, DATA, or META categories and generalizes specifics into {{placeholders}}. - Format Drift Guard: Runs a grep check on the target PROMPTS.md before writing to confirm the P-YYYYMMDD-### catalog format, preventing writes into the wrong file. - Use Case: After a productive debugging session, invoke the skill to extract the constraint-specification prompt that worked, templatize it, and append it as a P-ID entry in your DecisionLog prompt catalog. ## Quick Start Ask Claude to run the prompt-library skill to extract and save the effective prompts from this session.
