What problem does it solve? Research documents arrive in formats like PDF, DOCX, XLSX, and web captures that cannot be cited or computed on directly, and naive conversion silently mangles tables and equations. This Skill converts them into Markdown or structured artifacts while explicitly flagging uncertain OCR regions and keeping converted text clearly separated from verified sources. ## Core Features & Use Cases - Source-Type Routing: Identifies whether a document is an academic paper, survey codebook, official report, reviewer attachment, or web capture, then routes it to the correct downstream workflow such as literature verification or wave-crosswalk checks. - Parser Selection and Conversion: Chooses the lightest suitable parser (simple conversion, layout-aware extraction, table extraction, or manual inspection) and preserves page and section anchors for later citation. - Provenance and Caveat Recording: Produces a required output record capturing source path, parser used, conversion command, layout caveats, and verification status so extracted numbers are treated as unverified until checked. - Use Case: A researcher receives a survey codebook PDF whose table will drive a variable recode. The Skill extracts the table, flags uncertain cells, records the conversion caveats, and routes the coding claim through the wave-crosswalk check before use. ## Quick Start Convert the attached codebook PDF into a Markdown artifact with page anchors, flag any uncertain table cells, and record the parser and caveats in the output header.
