Generate premium enterprise-style PDFs from markdown content, with automatic de-AI text humanization (removes em dashes, AI filler phrases, overly formal language) and a professional Navy + Gold design system.
Generate high-quality, enterprise-grade PDF documents from markdown input. This skill applies a de-AI humanization pipeline to make LLM-generated text sound more natural, then renders it into a premium PDF with a professional Navy + Gold design system (marketing agency quality).
Invoke this skill when the user:
python3 premium-pdf/generate_pdf.py \
--input "# Report Title
Your markdown content here..." \
--output output.pdf \
--title "Document Title"
python3 premium-pdf/generate_pdf.py \
--input path/to/document.md \
--output output.pdf \
--title "Document Title"
python3 premium-pdf/generate_pdf.py \
--input path/to/document.md \
--output output.pdf \
--title "Quarterly Report" \
--logo path/to/company-logo.png
| Parameter | Required | Description |
|---|---|---|
--input or -i | Yes | Markdown text (string) or path to a .md file |
--output or -o | Yes | Output PDF file path (e.g., report.pdf) |
--title or -t | No | Document title shown in the header (default: "Enterprise Report") |
--logo | No | Path to a PNG or JPG logo image to display in the header |
Before rendering, the skill automatically applies text transformations:
—) → replaced with , (natural comma-based flow)–) → replaced with -The generated PDF features:
#, ##, ###, ####code and fenced code blocks-, *, +) and ordered lists (1., 2.)>)---)Install Python dependencies:
pip3 install -r premium-pdf/requirements.txt
Or install directly:
pip3 install "reportlab>=4.0.0" "Pillow>=9.0.0"
User: "Create a PDF from this markdown report and save it as quarterly-report.pdf"
Agent action:
python3 premium-pdf/generate_pdf.py \
--input "[user's markdown content]" \
--output quarterly-report.pdf \
--title "Quarterly Report Q1 2026"
Output: PDF generated successfully: /absolute/path/to/quarterly-report.pdf
After running, inform the user of the output path so they can open the file.
If the script returns an error:
reportlab is installed: pip3 install reportlab--logo is used)