What problem does it solve? Texts about AI often collapse critical distinctions—treating generated output as knowledge, fluency as understanding, or authority as evidence. This Skill provides a rigorous analytical framework to evaluate how a text handles the transition from production to evidence to attribution, preventing conceptual errors in epistemic analysis of AI-generated content. ## Core Features & Use Cases - Epistemic Distinction Analysis: Separates production, information, evidence, verification, attribution, authority, and truth claims within any text about AI. - Counterfactual Testing: Applies structured tests (remove generation, remove verification, replace AI with human production) to determine whether an issue is AI-specific or a general epistemic problem. - Structured Output Protocol: Produces findings with claim, evidence, verification, attribution, knowledge status, classification (A–E), confidence, and recommendation fields. - Use Case: Given a paper claiming "AI produces scientific knowledge," use this Skill to reconstruct what is actually produced, what evidential standards apply, how attribution is handled, and whether the claim conflates functional performance with understanding. ## Quick Start Analyze this text about AI-generated research summaries using the knowledge-evidence-production framework and report findings with the output protocol.
