PDF和图片文字提取

The skill claims to handle both images and PDFs. The included script (scripts/pdf_text_extractor.py) only processes text-based PDF content via PyMuPDF; image OCR is delegated in SKILL.md to a platform 'read_image' tool. SKILL.md also sometimes implies support for scanned PDFs but the script itself does not perform OCR — it will return empty text for image-only pages and advise using OCR. This is a minor mismatch but explainable (PDF handling vs OCR delegation).

Runtime instructions are narrow: obtain image URL or local PDF path, call the platform 'read_image' tool for images, and run the bundled Python script for PDFs. The script reads only the provided file path, validates suffix, extracts text with PyMuPDF, returns JSON, and does not access other files, environment variables, or external endpoints.

No automated install spec is included (instruction-only). SKILL.md documents a single dependency (pymupdf) and suggests 'pip install pymupdf>=1.23.0'. There are no downloads from arbitrary URLs or archive extraction steps in the package.

The skill requests no environment variables, credentials, or config paths. The code does not reference any secrets or external services. This is proportionate to the stated function.

The skill is not always-enabled, is user-invocable, and does not modify other skills or system settings. The script does not write persistent state; SKILL.md states files are only used in-session.