varunr89/claude-marketplace/pdf-ocr

PDF OCR

Converts PDF pages to images using PyMuPDF, then runs Apple Vision OCR on each page in parallel. Produces Markdown, plain text, or JSONL output.

When to use

Use this skill when the user wants to extract text from a PDF file -- especially scanned PDFs, image-based PDFs, or PDFs where copy-paste produces garbled text.

Usage

python3 ${CLAUDE_PLUGIN_ROOT}/scripts/pdf_ocr.py <input.pdf> \
  [-o output_file] \
  [-f markdown|text|jsonl] \
  [--dpi 200] \
  [--workers N] \
  [--languages en-US] \
  [--fast] \
  [--keep-images] [--images-dir <dir>] \
  [--stdout]

Key arguments

Pipeline

1. PDF pages are rendered to PNG images at the specified DPI using PyMuPDF (fitz)
2. Apple Vision OCR runs in parallel across pages using ThreadPoolExecutor
3. Results are assembled in page order and written to the chosen output format

Output formats

• markdown -- Each page becomes a ## Page N section with the OCR text
• text -- Plain text with === Page N === separators
• jsonl -- One JSON object per line with page, text, and backend fields

Dependencies

• PyMuPDF (pip install pymupdf) -- PDF to image conversion
• macOS + PyObjC (pip install pyobjc-core pyobjc-framework-Vision pyobjc-framework-Cocoa) -- Vision OCR