botrun-mcp-js-pdf

JavaScript PDF processing MCP Server - Pure JS implementation using pdf-parse

Features

This is a pure JavaScript implementation of PDF MCP Server, providing the same functionality as the Rust version (botrun-mcp-rust-pdf) but with:

• Zero native dependencies - Works on any platform with Node.js 18+
• Smaller package size - No binary downloads needed
• Easy to extend - Pure JavaScript, easy to customize

Available Tools

| Tool | Description | |------|-------------| | get_pdf_metadata | Get PDF metadata (page count, file size, token estimation) | | extract_pdf_text | Extract text with page range and character limit | | extract_pdf_chunked | Extract PDF in chunks for large files | | search_pdf_keyword | Search keywords with context (like ripgrep) |

Installation

For Claude Code

claude mcp add botrun-mcp-js-pdf -- npx botrun-mcp-js-pdf

For Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "js-pdf": {
      "command": "npx",
      "args": ["botrun-mcp-js-pdf"]
    }
  }
}

Global Installation

npm install -g botrun-mcp-js-pdf

Usage Examples

1. Get PDF Metadata (Always Start Here)

Use get_pdf_metadata on https://example.com/document.pdf

Returns:

• Total pages
• File size (MB)
• Estimated tokens
• AI strategy suggestion (extract vs chunked)

2. Extract Text from Specific Pages

Extract pages 1-5 from https://example.com/document.pdf

3. Extract Large PDF in Chunks

Get chunk 0 from https://example.com/large-document.pdf

4. Search for Keywords

Search for "introduction" in https://example.com/document.pdf

Local File Support

Both remote URLs and local files are supported:

# Remote URL
https://example.com/file.pdf

# Local file (only ./ prefix allowed for security)
./data/sample.pdf

Comparison with Rust Version

| Feature | JS Version (this) | Rust Version | |---------|------------------|--------------| | Performance | Good | Excellent | | Package Size | ~2MB | ~15MB | | Platform Support | Any (Node.js) | macOS, Linux | | Dependencies | Pure JS | Binary | | Customization | Easy | Requires Rust |

Choose JS version if:

• You want smaller package size
• You need to customize the code
• You're on a platform without pre-built binaries

Choose Rust version if:

• You need maximum performance
• You process many large PDFs
• You want the fastest extraction speed

Development

# Clone the repository
git clone https://github.com/bohachu/botrun-mcp-8.git
cd botrun-mcp-8/npm-packages/js-pdf-mcp

# Install dependencies
npm install

# Run tests
npm test

# Start server directly
npm start

Technical Details

• Built with @modelcontextprotocol/sdk
• PDF parsing by pdf-parse
• ESModule (type: module)
• Node.js 18+ required

License

MIT

Author

bohachu