Summary
- Automatically distill thinking frameworks from names, topics, or vague needs into executable AI agent skills.
- Supports two entry modes: direct name input for immediate distillation, or diagnostic questioning to recommend the best thinking framework for your specific problem Executes parallel six-agent research across writings, conversations, expression patterns, external views, decisions, and timeline to extract authentic mental models rather than surface-level traits Generates runnable Agentic Protocols so the skill researches facts before answering, avoiding hallucination on questions requiring real-world information Includes quality gates at research review, synthesis confirmation, and verification phases to ensure the final skill captures genuine thinking patterns, not generic AI platitudes