Convert PDFs to auditable JSONL with text extraction or LLM vision OCR.
npx skills add https://github.com/Jia-Hong-Peng/rag-to-skill --skill pdf-ocr-jia-hong-peng
Please help me install this Agent Skill. Skill: pdf-ocr Source: https://github.com/Jia-Hong-Peng/rag-to-skill/tree/main/pdf-ocr-skill Command: npx skills add https://github.com/Jia-Hong-Peng/rag-to-skill --skill pdf-ocr-jia-hong-peng
The skill automates turning PDFs into auditable, per-page JSONL records by smartly choosing between direct text extraction and LLM-based vision OCR, ensuring compliance with hard rules and guardrails.
Process a PDF to produce per-page JSONL records with chapter metadata, selecting text extraction for text PDFs or LLM Vision for scanned PDFs.
High-intent search queries and answers about installing and using this skill.
To convert a scanned PDF to JSONL, this skill renders image-based pages as PNGs and applies LLM vision OCR to transcribe text, appending per-page records with chapter metadata for structured output.
For PDFs with a text layer, the best way to extract text into JSONL is using PyMuPDF for direct text extraction, which produces stable, token-free outputs without requiring LLM vision OCR.
Yes, this PDF to JSONL conversion tool supports batch processing and resume capabilities, utilizing strict progress tracking and manifest generation to manage large conversion workflows.
Yes, you can extract chapter metadata and per-page records from a single book PDF by applying hard-rule guards that enforce single-book locking during the JSONL conversion process.
PyMuPDF is limited to extracting embedded text layers and cannot process scanned or image-based PDF pages, requiring the LLM vision OCR path to handle those files for successful JSONL conversion.
Guardrails like single-book locking are needed for PDF OCR to enforce hard rules, ensuring secure archival workflows, preventing concurrent processing conflicts, and maintaining manifest integrity.
Strategic-AutomationExtract text from PDFs and scanned documents using pymupdf and marker-pdf.
goodoleusaExtract structured data from screen recordings and UI screenshots into merged JSON.
lttrExtract text from PDFs and save clean Markdown beside the source file.
claude-office-skillsExtract text from images and scanned documents using PaddleOCR.
