PDF Form Filler

Fill static PDF forms by detecting PDF type and using the appropriate method:

• Text-based PDFs: Search for label text and insert values at found positions
• Image-based PDFs: Use Claude Vision iteratively for pixel-perfect placement

Quick Start

# Option 1: Text-based PDF (fast, uses label search)
uv run scripts/fill_pdf.py form.pdf filled.pdf "Label:?=value"

# Option 2: VLM-guided (works for ANY PDF, iteratively verifies placement)
uv run scripts/fill_pdf_vlm.py form.pdf filled.pdf --fields '{"Field": "value"}'

Method 1: Text Search Positioning (Text-Based PDFs)