pdf_mcp_server

This MCP server provides a set of PDF processing tools accessible via the FastMCP API. It exposes operations such as server_info to retrieve server metadata, list_temp_resources to inspect temporary files, get_pdf_info to read basic PDF properties, and upload-related helpers to store files in the server's temp directory. Tools return structured data with a meta block containing operation_id and execution_ms, and many endpoints wrap results in a consistent response shape. To interact programmatically, you can spawn the server with Python using -m fastmcp_pdf_server and then invoke the exposed tools through the MCP client, or use the provided programmatic usage example to call server_info and similar tools. The server is designed to operate over a STDIO transport; avoid printing extra output to stdout/stderr, and ensure logs are written to the configured log file as the server expects stdout to remain clean for MCP messaging.

Prerequisites:

• Python 3.8+ installed on the host
• pip available
• A POSIX or Windows environment with a writable temp directory

1. Create and activate a virtual environment (optional but recommended):

   • Windows PowerShell: python -m venv .venv ..venv\Scripts\Activate.ps1
   • Linux/macOS: python3 -m venv .venv source .venv/bin/activate

2. Install Python dependencies from requirements.txt: pip install -r requirements.txt

3. Configure environment variables (example values shown in mcp_config). Create or adjust the .env or environment as needed:

   • MAX_FILE_SIZE_MB=50
   • TEMP_DIR=/path/to/mcp_pdf_server/temp_files
   • LOG_LEVEL=DEBUG
   • LOG_FILE_PATH=/path/to/mcp_pdf_server/logs/fastmcp_pdf_server.log
   • SERVER_NAME=pdf-processor-server
   • SERVER_VERSION=1.0.0
   • PATH to include any required bin directories (e.g., if you rely on poppler or other tools)

4. Run the server (from project root): python -m fastmcp_pdf_server

5. If you installed the package system-wide or as a package, you can also start it with: fastmcp-pdf-server

Tips and common notes:

• The server uses STDIO transport. Do not print extra output to stdout or stderr; logs should be written to the configured log file.
• When dependencies are updated (for example, adding new packages like requests for URL uploads), reinstall dependencies with pip install -r requirements.txt.
• If you need to fetch file contents via MCP, use get_resource_base64 to obtain base64-encoded data from the temp directory, which is useful for clients without direct file transfer.
• Ensure TEMP_DIR exists and is writable before starting the server to avoid runtime errors.
• The server exposes REST-like tooling via the MCP interface; adapt your client to handle the meta information (operation_id and execution_ms) present in responses.