PDF Generation in Desktop Commander MCP: Libraries, Methods, and Code Examples

Desktop Commander MCP implements PDF generation through five specialized libraries: md-to-pdf for Markdown conversion, @puppeteer/browsers for Chrome management, @opendocsg/pdf2md for reverse conversion, unpdf for image extraction, and pdf-lib for structural editing.

Desktop Commander MCP is an open-source Model Context Protocol server that provides comprehensive document processing capabilities. The PDF generation implementation resides primarily within the src/tools/pdf/ directory, where specialized modules wrap industry-standard libraries to handle conversion, extraction, and manipulation tasks. Understanding these specific libraries and their integration patterns enables developers to extend the server's functionality or adapt similar architectures for their own TypeScript applications.

Core PDF Generation Libraries

The project adopts a modular approach, delegating specific PDF tasks to specialized libraries rather than relying on monolithic solutions.

Markdown to PDF Conversion with md-to-pdf

The primary engine for converting Markdown content to PDF documents is the md-to-pdf library. In src/tools/pdf/markdown.ts, the parseMarkdownToPdf() function wraps the library's mdToPdf() method, injecting Chrome executable paths discovered through the @puppeteer/browsers integration.

This conversion requires a Chromium-based browser to render the HTML intermediate. The code handles Chrome path resolution through getChromePath(), installChrome(), and findSystemChrome() functions defined in the same file (lines 86-122), ensuring the PDF generation pipeline never stalls due to missing browser binaries.

Chrome Discovery via @puppeteer/browsers

PDF generation reliability depends on consistent Chrome availability. The @puppeteer/browsers module provides the underlying mechanism for locating system Chrome installations or downloading browser binaries when needed. The ensureChromeAvailable() function proactively manages this dependency, kicking off background downloads if neither cached nor system-installed browsers are detected.

PDF to Markdown Reverse Conversion

For extracting text content from existing PDFs, Desktop Commander MCP utilizes @opendocsg/pdf2md. The wrapper function parsePdfToMarkdown() in src/tools/pdf/markdown.ts loads PDF files into Uint8Array buffers and passes them to the pdf2md() function. This implementation in src/tools/pdf/lib/pdf2md.ts (lines 70-78) returns structured data including page content, metadata, and references to embedded images.

Image Extraction with unpdf

Raster asset extraction leverages the unpdf library. The extractImagesFromPdf() function in src/tools/pdf/extract-images.ts utilizes unpdf's extractImages() method to pull image data from PDF page streams. This capability supports workflows requiring visual content analysis or document reconstruction.

Structural PDF Editing via pdf-lib

Manipulating PDF structure—such as deleting, inserting, or reordering pages—relies on the pdf-lib library. The editPdf() function in src/tools/pdf/manipulations.ts demonstrates this integration:

• PDFDocument.load(): Parses existing PDF bytes
• pdfDoc.removePage(): Deletes specified pages by index
• pdfDoc.insertPage(): Adds new pages at specific positions
• pdfDoc.save(): Serializes modifications back to bytes

Implementation Examples

The following code patterns demonstrate how these libraries work together within the Desktop Commander MCP architecture.

Generate PDF from Markdown Content

import { parseMarkdownToPdf } from '@/tools/pdf/markdown.js';

const md = '# Hello World\nThis is a PDF generated from Markdown.';

const options = { 
  pdf_options: { 
    format: 'A4', 
    margin: { top: '1in' } 
  } 
};

const pdfBuffer = await parseMarkdownToPdf(md, options);
// Returns Node Buffer containing PDF file bytes

This calls mdToPdf() internally, automatically resolving Chrome paths through getChromePath().

Convert PDF to Structured Markdown

import { parsePdfToMarkdown } from '@/tools/pdf/markdown.js';

const result = await parsePdfToMarkdown('docs/example.pdf');

console.log(result.metadata.title);
console.log(result.pages[0].text);

The function delegates to pdf2md() from @opendocsg/pdf2md after converting the file to Uint8Array.

Edit PDF Structure

import { editPdf } from '@/tools/pdf/manipulations.js';

const operations = [
  {
    type: 'delete',
    pageIndexes: [1, 2]  // zero-based indexes
  },
  {
    type: 'insert',
    pageIndex: 0,
    markdown: '# New Intro\nAdded via md-to-pdf.'

  }
];

const editedPdf = await editPdf('docs/example.pdf', operations);
// Returns Uint8Array of modified PDF

The editPdf() function uses pdf-lib to load the source document, apply the operation schema, and generate modified bytes.

Ensure Chrome Availability at Startup

import { ensureChromeAvailable } from '@/tools/pdf/markdown.js';

// Call once when server initializes
ensureChromeAvailable();

This preventive measure ensures PDF generation capabilities remain available without runtime interruptions.

File Structure and Architecture

The PDF generation capabilities are organized across specific tool modules:

• src/tools/pdf/markdown.ts: Core PDF generation utilities; wraps md-to-pdf, Chrome discovery, and PDF-to-Markdown conversion
• src/tools/pdf/lib/pdf2md.ts: Thin wrapper around @opendocsg/pdf2md for parsing logic
• src/tools/pdf/manipulations.ts: PDF editing operations built on pdf-lib
• src/tools/pdf/extract-images.ts: Image extraction logic using unpdf
• src/utils/files/pdf.ts: High-level façade exposing PDF operations to the broader application

The package.json file declares these external dependencies: md-to-pdf, @opendocsg/pdf2md, pdf-lib, unpdf, and @puppeteer/browsers.

Summary

• Desktop Commander MCP implements PDF generation through five specialized npm packages rather than custom engines
• md-to-pdf handles Markdown-to-PDF conversion with Chrome/Chromium rendering via @puppeteer/browsers
• @opendocsg/pdf2md provides reverse conversion capabilities from PDF to structured Markdown
• unpdf extracts raster images from PDF documents for content analysis
• pdf-lib enables structural modifications including page deletion, insertion, and layout changes
• All functionality resides under src/tools/pdf/ with clear separation between conversion, extraction, and manipulation concerns

Frequently Asked Questions

What is the primary library for Markdown to PDF conversion in Desktop Commander MCP?

The md-to-pdf library serves as the primary engine for converting Markdown content to PDF documents. The parseMarkdownToPdf() function in src/tools/pdf/markdown.ts wraps this library, handling both the conversion logic and Chrome browser dependency injection through the launch_options parameter.

How does Desktop Commander MCP handle Chrome browser dependencies?

The server utilizes @puppeteer/browsers to manage Chrome availability through three main functions: findSystemChrome() checks for existing installations, installChrome() downloads browser binaries when needed, and getChromePath() returns the executable path. The ensureChromeAvailable() function proactively ensures browser readiness before PDF generation requests occur.

Can Desktop Commander MCP extract images from existing PDF files?

Yes, the unpdf library provides image extraction capabilities through the extractImages() method. This functionality is implemented in src/tools/pdf/extract-images.ts and enables the server to pull raster assets from PDF page streams for further processing or analysis.

Which library handles structural modifications like page deletion and insertion?

pdf-lib manages all structural PDF editing operations. The editPdf() function in src/tools/pdf/manipulations.ts uses PDFDocument.load() to parse existing documents, applies operations through methods like removePage() and insertPage(), and serializes results via pdfDoc.save() to produce modified PDF bytes.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}