ComPDF Skills — AI Agent–Ready PDF Processing Toolkit

As part of the KDAN ecosystem, ComPDF Skills work with 39+ AI coding agents including Claude Code, Cursor, GitHub Copilot, OpenCode, Windsurf, Gemini CLI, and more.

Get started in seconds — no complex setup, just npx skills add and you're ready to process PDFs through your AI agent.

⭐ Star this repo if you find it helpful! Questions or feedback? Join our Discussions.

Why ComPDF Skills?

• Works Everywhere — Compatible with 39+ AI coding agents and integrates with Salesforce, SharePoint, Make, Zapier, and MCP tools
• 10+ Output Formats — Word, Excel, PPT, HTML, Markdown, JSON, CSV, RTF, TXT, Image — all from a single skill
• Local & Cloud Options — Choose local processing for sensitive documents or cloud API for zero-install convenience
• Free to Start — 200 free conversions (SDK trial) or 200+ free monthly API calls
• Enterprise-Grade — Secure infrastructure with data privacy compliance

Requirements

Free Trial & License

Each ComPDF Skill uses a different license mechanism. See the table below for details and the individual skill sections for step-by-step guidance. If you need to upgrade or purchase, contact our sales.

Provided Skills

pdf-to-word-docx — Convert PDFs/images to 10 Formats

This is a local skill for image and PDF conversion. Converts PDFs and images to Word, Excel, PPT, HTML, Markdown, JSON, CSV, RTF, TXT, and Image with AI-powered layout analysis and OCR.

Usage examples:

# PDF to Word (default: AI layout analysis enabled)
npx skills run pdf-to-word-docx word input.pdf output.docx

# PDF to Excel, one worksheet per page
npx skills run pdf-to-word-docx excel input.pdf output.xlsx --excel-worksheet-option for-page

# PDF to Markdown (great for LLM data prep)
npx skills run pdf-to-word-docx markdown input.pdf output.md

# Scanned image to Word (OCR auto-enabled)
npx skills run pdf-to-word-docx word scan.png output.docx --ocr-language english

# PDF to HTML, multi-page with bookmarks
npx skills run pdf-to-word-docx html input.pdf output_dir --html-option multiple-page-with-bookmark

# PDF with specific page range
npx skills run pdf-to-word-docx word input.pdf output.docx --page-ranges "1-3,5"

When to use: Batch document archiving, content extraction for LLM input, invoice/contract digitization, scanned document OCR.

Trial: 200 free conversions with auto-downloaded license.xml from download.compdf.com. AI model (~525MB) downloads on first run. To upgrade, replace scripts/license.xml with your purchased license file — no conversion limits.

pdf-editor-compdf — Full PDF Page Editing

The pdf-editor-compdf is designed to edit, compare, and compress PDFs. All processing is 100% local — no file leaves your machine.

Usage examples:

# Split a PDF into individual pages
npx skills run pdf-editor-compdf split "report.pdf" --mode all --overwrite

# Merge multiple PDFs
npx skills run pdf-editor-compdf merge "a.pdf" "b.pdf" --output "merged.pdf" --overwrite

# Extract pages 2-5
npx skills run pdf-editor-compdf extract "input.pdf" --range 2-5 --output "out.pdf" --overwrite

# Convert to PDF/A-1a for archiving
npx skills run pdf-editor-compdf convert "report.pdf" --standard pdfa-1a --overwrite

# Compress images in a PDF
npx skills run pdf-editor-compdf optimize "report.pdf" --compress-images --image-quality 50 --overwrite

# Add "CONFIDENTIAL" watermark
npx skills run pdf-editor-compdf watermark-text "report.pdf" --text "CONFIDENTIAL" --overwrite

# Compare two PDF versions visually
npx skills run pdf-editor-compdf compare "old.pdf" "new.pdf" --overwrite

When to use: Contract splitting/merging, version comparison, watermarking, PDF/A compliance for legal/regulatory archives, file size optimization.

Trial: 30-day free SDK license — on first use, provide your email; the skill sends an activation request to wms.compdf.com/api/license/skillsTrial and stores the license in scripts/*/license_key_*.xml. Run --help to confirm activation. CLI binary auto-downloads. To upgrade, Contact Sales.

pdf-tools-compdf — 50+ Cloud PDF Operations

Process PDFs through ComPDF Cloud REST API. Convert formats, edit pages, apply OCR, extract text and tables, and more — no local dependencies required. Works on any platform.

Usage examples:

# PDF to Word (cloud)
npx skills run pdf-tools-compdf --convert pdf/docx --file input.pdf

# PDF to Image
npx skills run pdf-tools-compdf --convert pdf/img --file input.pdf

# Merge PDFs
npx skills run pdf-tools-compdf --convert pdf/merge --file "file1.pdf,file2.pdf"

# OCR a scanned document
npx skills run pdf-tools-compdf --convert documentAI/ocr --file scan.pdf

# Extract structured data with AI
npx skills run pdf-tools-compdf --convert idp/documentExtract --file invoice.pdf

# Compress PDF
npx skills run pdf-tools-compdf --convert pdf/compress --file input.pdf

When to use: Lightweight cloud integration, no-install environments, batch processing via API, AI-powered document extraction and parsing.

Trial: 200+ free API calls per month. Register at ComPDF Cloud dashboard to get your API Public Key. Authentication is via x-api-key HTTP header. The key can be stored in config/public_key.txt (optional, user-consented). No local license file required.

Install

ComPDF Skills are available through multiple installation channels. Choose the one that fits your workflow.

Via skills.sh (Recommended)

Install all skills at once with a single command — works across 20+ AI coding agents:

npx skills add ComPDFKit/compdf-skills

Install a specific skill:

npx skills add ComPDFKit/compdf-skills@pdf-to-word-docx
npx skills add ComPDFKit/compdf-skills@pdf-editor-compdf
npx skills add ComPDFKit/compdf-skills@pdf-tools-compdf

Via ClawHub

Browse and install individual skills directly from ClawHub:

• pdf-to-word-docx
• pdf-editor-compdf
• pdf-tools-compdf

Via MCP (Model Context Protocol)

For MCP-compatible agents, use the ComPDF MCP servers:

• MCP.SO — ComPDF MCP Server
• MCPMarket — ComPDF MCP Server

Compatible Agents

The npx skills add command works out of the box with these AI coding agents: