What problem does it solve? Studying dense PDFs, textbooks, and papers is slow and passive; this Skill turns documents into an interactive tutoring environment with a searchable knowledge base, entity-relationship knowledge graph, step-by-step problem solving, and practice question generation. ## Core Features & Use Cases - Knowledge Base Initialization: Chunks documents and builds a NetworkX knowledge graph of concepts, definitions, theorems, and their relationships. - Dual-Loop Problem Solving: Separates an Analysis Loop (targeted hybrid retrieval with BM25 plus graph expansion) from a Solve Loop (step-by-step reasoning with citations). - Question Generation & Guided Learning: Creates difficulty-calibrated practice questions grounded in document content and teaches topics along prerequisite-aware learning paths. - Deep Research Reports: Decomposes a topic into subtopics and produces multi-section cited academic reports. - Use Case: Attach a machine learning textbook, ask to build a knowledge base, then request "quiz me on backpropagation at medium difficulty" or "explain gradient descent step by step with citations." ## Quick Start Ask Claude to use deeptutor to create a knowledge base from your attached PDF and then explain a concept from it step by step.
