PDF MCP Server MCP Server for AI agents to read, write, manipulate, and create PDF documents. Provides deterministic, stateless PDF operations with structured outputs. ## Features - Extract text from PDF documents (no OCR) - Get metadata including title, author, page count - Fill form fields in interactive PDFs - Merge multiple PDFs into one document - Split PDFs into separate files by page ranges - Create PDFs from HTML with full CSS support - Render pages to images (PNG/JPEG) ## How to Use ### Step 1: Select a Tool Choose from 11 available PDF operations: | Tool | What it does | |------|-------------| | pdf.open | Load PDF from URL or base64 | | pdf.get_metadata | Get document info (title, author, pages) | | pdf.extract_text | Extract all text content | | pdf.render_pages | Convert pages to PNG images | | pdf.list_form_fields | List all fillable form fields | | pdf.fill_form_fields | Fill form fields with values | | pdf.flatten_form | Make form fields permanent | | pdf.create_from_html | Generate PDF from HTML content | | pdf.merge | Combine multiple PDFs | | pdf.split | Split PDF by page ranges | | pdf.extract_pages | Extract specific pages | ### Step 2: Provide Input Enter the PDF URL in the PDF URL field, or provide base64-encoded content. ### Step 3: Run Click Start to execute. Results appear in the dataset, artifacts in the key-value store. ## Input Examples ### Get Metadata json { "tool": "pdf.get_metadata", "pdfUrl": "https://example.com/document.pdf" } ### Extract Text json { "tool": "pdf.extract_text", "pdfUrl": "https://example.com/document.pdf", "pages": "1,2,3" } ### Create from HTML json { "tool": "pdf.create_from_html", "html": "<h1>Hello World</h1><p>This is a PDF.</p>" } ### Merge PDFs json { "tool": "pdf.merge", "urls": [ "https://example.com/doc1.pdf", "https://example.com/doc2.pdf" ] } ### Split PDF json { "tool": "pdf.split", "pdfUrl": "https://example.com/document.pdf", "ranges": "[{\"start\":1,\"end\":5},{\"start\":6,\"end\":10}]" } ## Output Format All tools return a consistent structure: json { "type": "tool_result", "tool": "pdf.extract_text", "status": "ok", "errors": [], "warnings": [], "artifacts": [] } ### Artifacts Generated files (PDFs, images) are stored in the run's key-value store with direct download URLs. ## Use Cases - Document Processing: Extract text for analysis or indexing - Form Automation: Fill tax forms, applications, contracts programmatically - Report Generation: Create PDFs from HTML templates - Document Assembly: Merge multiple PDFs into packages - Page Extraction: Pull specific pages from large documents ## Limits | Limit | Value | |-------|-------| | Max file size | 50 MB | | Max pages | 500 | | Render DPI | 150 | ## Integration This actor implements the MCP (Model Context Protocol) standard for AI agent integration. Use it with Claude, GPT, or any MCP-compatible agent. ## Support For issues or feature requests, contact the author. --- PDF MCP Server | Signature: [Claude-Jaclyn]