Agent Skills: Document Exporter Skill

Convert Markdown documents into PDF, DOCX, PPTX, or XLSX formats. Routes Office formats to OfficeCLI, PDF via Chrome.

UncategorizedID: officebeats/beats-pm-antigravity-brain/document-exporter

Install this agent skill to your local

pnpm dlx add-skill https://github.com/officebeats/beats-pm-kit/tree/HEAD/.agent/skills/document-exporter

Skill Files

Browse the full folder contents for document-exporter.

Download Skill

Loading file tree…

.agent/skills/document-exporter/SKILL.md

Skill Metadata

Name
document-exporter
Description
Convert Markdown documents into PDF, DOCX, PPTX, or XLSX formats. Routes Office formats to OfficeCLI, PDF via Chrome.

Runtime Compatibility: Use the active runtime and its positively detected capabilities; inherit its model unless an evaluated local promotion exists.

Document Exporter Skill

Use this skill when the user requests a PDF, Word Doc, PowerPoint, Excel export, or "Export" of a markdown file.

Capabilities

  1. Markdown → HTML → PDF: Creates a self-contained HTML file, then uses Headless Chrome to print to PDF.
  2. Markdown → DOCX (via OfficeCLI): Creates a real Word document with proper formatting.
  3. Markdown → PPTX (via OfficeCLI): Creates a PowerPoint presentation from structured content.
  4. Markdown → XLSX (via OfficeCLI): Creates an Excel workbook from tabular data.

Format Selection

| User Says | Action | |-----------|--------| | "PDF", "print" | Use existing HTML/PDF pipeline (Steps 1-4 below) | | "Word", "DOCX", "document" | Delegate to office-cli skill | | "PowerPoint", "PPTX", "deck", "presentation", "slides" | Delegate to office-cli skill | | "Excel", "XLSX", "spreadsheet", "tracker" | Delegate to office-cli skill | | "export" (no format specified) | Ask user; default PPTX for presentations, DOCX for docs, PDF for reports |

PDF Pipeline Instructions

1. Locate the Markdown File

Ensure you have the absolute path to the .md file.

2. Generate HTML (Intermediate Step)

Run the md_to_pdf.py script located in scripts/.

python3 .agent/skills/document-exporter/scripts/md_to_pdf.py [INPUT_MD] [OUTPUT_HTML]

3. Convert to PDF (Final Step)

Use Headless Chrome to convert the HTML to PDF.

# macOS
"/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" --headless --disable-gpu --print-to-pdf="[OUTPUT_PDF]" "[OUTPUT_HTML]"

# Windows
& "C:\Program Files\Google\Chrome\Application\chrome.exe" --headless --disable-gpu --print-to-pdf="[OUTPUT_PDF]" "[OUTPUT_HTML]"

4. Verify & Notify

  • Check the file size of the PDF (should be > 100KB if images are included).
  • Notify user with the file path.

Office Format Instructions

For DOCX, PPTX, or XLSX exports, load and follow the office-cli skill:

.agent/skills/office-cli/SKILL.md

The office-cli skill has complete workflows for:

  • Creating PowerPoint decks with slides, shapes, tables, charts, and images
  • Creating Word documents with paragraphs, styles, tables, and images
  • Creating Excel workbooks with data, formulas, charts, and pivot tables

Quick Delegation Example

# User wants a PowerPoint from a markdown outline:
# 1. Parse the markdown headings as slide titles
# 2. Parse bullet points as slide content
# 3. Use officecli batch mode for performance

officecli create output.pptx
officecli batch output.pptx --commands '[
  {"command":"add","parent":"/","type":"slide","props":{"title":"Slide Title from H1"}},
  {"command":"add","parent":"/slide[1]","type":"shape","props":{"text":"Content from bullets","x":"2cm","y":"5cm","w":"22cm","h":"12cm","font":"Arial","size":"18"}}
]' --json