Extract text, search inside PDFs, and produce summaries.
Fixed SKILL.md to match actual PyMuPDF implementation. Corrected API documentation.