Universal PDF Master

Problem

PDF workflows often fail because the file state is unclear before processing: encryption, scanned pages, missing text layers, or mixed content can cause extraction, merging, compression, and redaction steps to fail silently. This skill makes detection and routing explicit before any transformation.

How It Works

• Detect first: pdf_info.py reads page count, size, encryption status, text layer, scan ratio, and metadata.
• Route by type: text-based files use PyMuPDF / pdfplumber; image-only files use OCR; mixed pages are processed by page; encrypted files require a password or stop.
• Split intent: compound requests become ordered capability chains, such as rename, merge, and index generation, with progress after each step.
• Confirm risky operations: batch jobs, irreversible changes, and missing parameters trigger a checklist, diff, or default-value confirmation.
• Deliver with QA: output includes the result file, acceptance findings, error fallbacks, and security notes; invoice verification and e-signature are explicitly limited when external services are unavailable.

Boundaries

It suits controlled PDF editing, organization, extraction, and redaction workflows. Confidential files must follow local-processing constraints. Black-out/redaction and permanent deletion are distinct operations and must not be conflated. If external dependencies are unavailable, it must not fabricate verified invoices or legally effective signatures.