Verifies user comprehension of AI-generated work through structured Socratic probing and gap analysis.
Grasp (Katalepsis) is an epistemic protocol designed to bridge the gap between AI execution and human comprehension. When an AI produces complex code or structural changes that are difficult to follow, this skill initiates a structured dialogue rather than providing a generic explanation. It categorizes the AI's work into specific domains like architecture or logic, identifies precise comprehension gaps such as causality or scope, and uses Socratic questioning to ensure the user truly understands the results. This transforms 'ungrasped results' into 'verified understanding,' ensuring the human operator remains firmly in control of the codebase.
Key Features
01Convergence evidence that demonstrates understanding through user reasoning
02Automatic categorization of AI work into architectural, logical, or dependency domains
03Multi-phase gap detection identifying issues in causality, scope, and sequence
04Progress tracking via internal task registration and updates
05Structured Socratic probing to verify human comprehension
0688 GitHub stars
Use Cases
01Reviewing complex architectural refactors where the rationale is not immediately obvious
02Debugging subtle logic changes where the user needs to grasp the specific sequence of operations
03Onboarding onto a codebase after the AI has generated significant new functionality
