视觉画布设计
canvas-design
基于设计原则创作静态视觉作品,用于海报与品牌图形。
document-core【文档与办公】Open, inspect, authenticate, convert, save, and safely manage PyMuPDF documents and shared geometry primitives.
独立 GitHub 快照 · 自动分类 · 以原作者文档为准
Use this sub-skill for the core PyMuPDF document lifecycle: open or create a document, inspect page/metadata/outline basics, authenticate protected PDFs, convert supported non-PDF inputs to PDF, choose a safe save mode, or reason about shared geometry primitives.
Route text/OCR/tables/RAG to ../text-tables-and-rag/SKILL.md, rendering/images to ../rendering-images-and-graphics/SKILL.md, PDF editing/forms/redaction to ../pdf-editing-annotations-forms/SKILL.md, and CLI/build/test concerns to ../cli-and-maintenance/SKILL.md.
Use import pymupdf; open local paths with pymupdf.open(path), memory with pymupdf.open(stream=data, filetype="pdf"), and text-like inputs with an explicit filetype. Authenticate when doc.needs_pass is true. Prefer full save to a new path unless intentional incremental update is supported. Reacquire pages and child objects after page-tree changes or document close/reopen.