Split, merge, and rename OCR-processed legal PDF documents by content.
npx skills add https://github.com/cat-xierluo/legal-skills --skill pdf-organizer
Please help me install this Agent Skill. Skill: pdf-organizer Source: https://github.com/cat-xierluo/legal-skills/tree/main/skills/pdf-organizer Command: npx skills add https://github.com/cat-xierluo/legal-skills --skill pdf-organizer
Scanned legal case files often arrive as one large merged PDF or as arbitrarily split fragments with meaningless filenames, forcing lawyers to manually identify document boundaries and rename files. This Skill organizes OCR-processed PDFs into properly named individual legal documents based on their actual content.
Ask the AI to inspect the scanned case PDF, generate an organize manifest draft, review the proposed document boundaries and filenames, then execute the split and rename into an output folder.
High-intent search queries and answers about installing and using this skill.
Run the script with --inspect to extract per-page titles, page numbers, and boundary signals, then generate a draft manifest with --suggest-manifest. After reviewing the manifest, execute it with pdf_organizer.py to split the PDF into individually named documents.
Define source_items in the organize manifest listing the fragment files in order, optionally with page ranges. The script merges them into one output PDF using pypdf and applies the suggested standardized filename.
No. The script runs in strict text-layer mode by default and stops if no searchable text is detected. You must first process the scan with an OCR tool to create a dual-layer PDF before content-based splitting.
Yes. Set rotate to 90, 180, or 270 degrees per segment or globally in the manifest. Deskew is available when ocrmypdf is installed; otherwise complex image correction should be handled by a dedicated PDF processor.
Low-confidence segments are marked needs_review with conservative filenames like "待确认 页码.pdf" rather than guessed names. The handoff.json flags review_required so downstream skills route these documents back for manual verification.
nadicodeaiExtract and convert PDF, scanned document, and image text to plain text or Markdown using OCR and pymupdf/marker-pdf.
AlsKozlovExtract predefined fields from Russian legal documents into risk-tagged tabular output.
davpatel605-beepExtract text, markdown, and metadata from PDFs and scanned documents.
erichugyEnforce descriptive file and folder naming conventions in repositories.
