Skill: pdf-extract
Goal
Extract exhibitor/company tables or contact blocks from PDFs (catalogs, lists).
Inputs
- •data/inputs/**/*.pdf
Outputs
- •staging/pdf_extract.parquet (row-level)
- •outputs/pdf_extract_report.md
Tools
- •pdfplumber first (text + table heuristics)
- •camelot only if needed and PDF is tabular (archived project risk)