What problem does it solve? Answering questions about scientific papers often requires combining information scattered across methods sections, figures, tables, supplements, and cited references, which is error-prone when done from a single visible page. ## Core Features & Use Cases - Structured PDF Parsing: Maps a paper's structure (title, abstract, methods, figures, tables, supplements, citations) using pdftotext or layout-aware tools so content sits on disk for reasoning. - Provenance-Tracked Extraction: Pulls exact pages and regions supporting an answer, keeping table values, figure labels, accession IDs, and quotes tied to page numbers. - Cross-Checking Claims: Verifies conclusions against methods, captions, supplement text, and cited papers. - Use Case: When reviewing a genomics paper, extract the data accession IDs from the methods section, confirm the figure captions match the claimed results, and save cited notes with page references under notes/ or docs/. ## Quick Start Use the pdf-explore skill to extract the methods, table values, and data accessions from the attached paper and cross-check the main claims against its figures and supplement.
