This PDF skill equips an agent to operate on PDF documents reliably: extract text and tables, merge or split files, rotate and watermark pages, and run OCR on scanned documents. It emphasizes reproducible scripted operations and verifies outputs (page counts, extraction quality) so downstream analysis or manuscript workflows remain auditable.
Use this skill whenever a user asks the agent to interact with PDF artifacts — for example: extracting tables from a research paper, merging multiple reports into a single deliverable, splitting large scans into chapters, applying watermarks for distribution, or OCRing scanned pages for searchable text. It’s suited to reproducible research pipelines and document-prep tasks.
Best for agents with Python runtime access and file I/O (e.g., Claude Code, Copilot/Code assistants, or any agent able to run pypdf scripts). Suitable for paired-research assistants and workflow automation agents.
This skill is a thin wrapper around Anthropic's official PDF skill, providing minimal instructions — just a brief SKILL.md with triggers, working rules, and a tiny pypdf code snippet. No scripts are bundled, no output contracts defined, and the actual implementation is left entirely to the agent. The skill is essentially a pointer to another repo's skill with a few bullet points.
Derived/forked from Anthropic's official skills repo. Minimal content — mostly a pointer with brief working rules. No security concerns since there's no executable code. Low usefulness because it provides almost no automation; the agent could follow the same rules without this skill.