Frequently Asked Questions (FAQ) - Pandoc MCP Server

Q: What is the Pandoc MCP Server? A: The Pandoc MCP Server is a Python-based tool that allows AI agents (like Claude) to convert documents between various formats (Markdown, DOCX, PDF, HTML, etc.) using Pandoc, the universal document converter. It uses the Model Context Protocol (MCP) for easy integration.

Q: What is MCP (Model Context Protocol)? A: MCP is an open protocol that standardizes how applications provide context to LLMs. The MCP Server acts as a bridge, allowing AI models to access and interact with external data sources and tools.

Q: What formats does the Pandoc MCP Server support? A: It supports a wide range of input and output formats handled by Pandoc, including Markdown, DOCX, PDF, HTML, EPUB, and many more. For a complete list, refer to the Pandoc documentation and look for the --list-output-formats option.

Q: How do I install the Pandoc MCP Server? A: You can install it either using Docker (recommended) or locally by manually installing the dependencies (Python, Pandoc, LaTeX). Detailed instructions are provided in the project’s README file.

Q: What are the benefits of using Docker for deployment? A: Docker simplifies installation and ensures consistent performance by bundling all dependencies into a container. It eliminates potential conflicts between different software versions on your system.

Q: How do I use the convert_document tool? A: The convert_document tool requires you to specify the input file path, output file path, and the desired output format. Optionally, you can also specify the input format and pass extra command-line arguments to Pandoc.

Q: What is the UBOS Asset Marketplace? A: The UBOS Asset Marketplace is a platform where you can find and deploy various tools and integrations for building and deploying AI agents. The Pandoc MCP Server is available on the UBOS Asset Marketplace for easy access and integration with the UBOS platform.

Q: How does the Pandoc MCP Server integrate with LangChain or LangGraph? A: The server is MCP compliant, making it compatible with MCP clients like LangChain’s MultiServerMCPClient. You can configure your client to connect to the server’s endpoint (e.g., http://127.0.0.1:8000/sse).

Q: What if I encounter an error during conversion? A: The server returns an error message indicating the cause of the error. Common errors include incorrect file paths, unsupported formats, or issues with the Pandoc installation. Check the server logs for more detailed information.

Q: How do I specify additional Pandoc arguments? A: You can use the extra_args parameter in the convert_document tool to pass additional command-line arguments directly to Pandoc. For example, you can add a table of contents to a PDF file using ['--toc'].

Q: What if I want to convert files located outside of the Docker container? A: You need to mount a volume to share a directory between your host machine and the Docker container. This allows the server to access files on your host machine by specifying the path inside the container (e.g., /data/my_file.docx).

Q: Do I need LaTeX installed for all conversions? A: No, LaTeX is only required for PDF generation. If you are not converting to PDF, you do not need to install LaTeX.

Q: How can I contribute to the project? A: Contributions are welcome! You can submit a Pull Request or open an Issue on the project’s GitHub repository.