PDF Read

Universal text reading (pdf-read)

Universal text reading

Part of the Universal PDF Management & Creation Skill Repository. Category: PDF Management. Every skill is individually callable (INDIVIDUAL MODE) and may also be invoked automatically by the orchestrator (SYSTEM MODE).

When to use

Pull readable text out of a digital PDF without loading it into the model.

Operations

• read full text
• read selected pages
• read preserving page boundaries
• stream huge documents page-by-page
• detect whether OCR needed

Engine (what actually executes this)

universal_pdf/core.extract:extract_text / extract_per_page / iter_page_text / detect_text_layer

All heavy lifting is done on disk / one page at a time by the shared engine — never by loading the document into an LLM context.

Tools & graceful degradation

• Preferred tool: PyMuPDF text layer
• Fallback: pdfplumber (layout text)

If the preferred backend is unavailable the skill transparently uses the fallback or reports a documented limitation. It never reports success when an operation did not actually happen.

Validation (this skill attaches checks to its Outcome)

text_coverage; page coverage

Related skills

pdf-extract, pdf-inspect, pdf-ocr, pdf-search, pdf-orchestrate

Example requests it handles

• "Read this PDF and tell me the main points."
• "Show me the text on pages 4 and 7."

Notes

Provenance: outputs cite their source page/section where relevant.