Document generation tools for creating professional presentations, reports, and formatted documents.
Create professional PowerPoint presentations using python-pptx.
Features:
- Title, section, content, and closing slide types
- Custom color schemes and typography
- Visual accents (bars, lines, backgrounds)
- Square bullet markers
- Images and tables support
Invocation:
ac-document-gen:powerpoint
Example:
"Create a presentation about our Q4 results with 5 slides"
Generate QR codes from URLs, text, or other data in multiple formats.
Features:
- SVG, PNG, EPS, and PDF output formats
- Custom colors, scale, and border
- Error correction levels (L/M/Q/H)
- Micro QR Code support
- Transparent backgrounds
Invocation:
ac-document-gen:qr-code
Example:
"Generate a QR code for https://example.com as an SVG"
Parse PDF files to markdown using GLM-OCR via Ollama locally.
Features:
- Local OCR via GLM-OCR model on Ollama (no cloud APIs)
- Page-by-page conversion with
<!-- Page N -->markers - Automatic dependency checking and guided installation
- Background processing for large PDFs (>10 pages)
- Bullet character normalization to standard markdown
Invocation:
ac-document-gen:pdf-ocr <path-to-pdf>
Example:
"Convert my report PDF to markdown"
Requirements: Ollama with glm-ocr model, poppler (pdftoppm/pdfinfo), sips (macOS built-in)
claude plugins install alteredcraft-plugins/ac-document-genClone and add to your Claude Code plugins:
git clone https://github.com/AlteredCraft/claude-code-plugins.git
claude plugins install --local ./claude-code-plugins/plugins/document-genEach skill follows a consistent workflow:
- Ask where to save the output file
- Gather requirements from the user
- Generate the output using Python scripts via
uv run - Deliver to the specified location
- powerpoint: Sets up a temp project in
/tmp/claude/pptx-project/, generates the.pptx, delivers it, and cleans up. - qr-code: Runs a single script with PEP 723 inline metadata — no temp project needed. Dependencies (
segno) are resolved automatically byuv run. - pdf-ocr: Checks local dependencies (Ollama, glm-ocr, poppler), converts each PDF page to an image, runs OCR via Ollama, and assembles the results into a single
.mdfile.
MIT