What problem does it solve? Reading PDFs—especially math lecture notes and academic papers—is hard for text-only tools because equations, diagrams, and figures get mangled by plain text extraction. This Skill combines text extraction with page rendering so an AI can both read the text and visually inspect math-heavy or figure-heavy pages. ## Core Features & Use Cases - PDF Triage: Run pdf_info.py to get page count, table of contents, per-page text length, image count, and math density before deciding how to read the document. - Hybrid Reading Strategy: Extract text for cheap structural overview, then render only math- or diagram-heavy pages to PNG images for visual comprehension at configurable DPI. - Targeted Search: Use pdf_search.py with regex or literal queries to locate theorems, definitions, or specific passages with surrounding context lines. - Use Case: Given a 40-page math paper, triage it with pdf_info.py, extract all text for structure, render only the high-math-density pages, and answer questions like "what does theorem 3.2 say" with precise LaTeX statements. ## Quick Start Ask the AI to read and summarize the attached PDF paper, rendering any pages that contain dense equations or figures.
