mcp-server-slides-converter

enesgnl/mcp-server-slides-converter

3.2

If you are the rightful owner of mcp-server-slides-converter and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

SlideConverter is an MCP server designed to convert PDF slide decks into a 4-up layout, providing the result as a Base64-encoded PDF.

Tools
1
Resources
0
Prompts
0

mcp-server-slides-converter

SlideConverter is an MCP server that converts a PDF slide deck into a 4-up layout (four slide thumbnails per page) and returns the result as a Base64-encoded PDF.

Requirements

  • Python 3.10+
  • No Poppler required (uses PyMuPDF/fitz for rendering)

Installation

pip install -r requirements.txt

Run locally

python slides_converter.py

MCP Integration

This repository exposes an MCP server discoverable via the mcp.json manifest over stdio.

  • Server command: python slides_converter.py
  • Tool signature: convert_pdf_4up(input_pdf_base64?: string, input_pdf_path?: string, dpi?: int = 72)
  • Output: { filename: string, pdf_base64: string, mime_type: "application/pdf" }

Run with uv

uv run -p 3.10 fastmcp mcp.json

If Python 3.10+ is not available, install 3.10/3.11 (e.g., with pyenv) and select it via the -p flag.

Notes

  • Inputs: provide either input_pdf_base64 (Base64-encoded PDF bytes) or input_pdf_path (absolute/working-directory file path).
  • dpi default is 72 for faster processing and smaller outputs. Increase for higher quality if needed.

Saving the returned PDF (client-side example)

import base64

result = tool_call_result  # { filename, pdf_base64, mime_type }
with open(result["filename"], "wb") as f:
    f.write(base64.b64decode(result["pdf_base64"]))

Packaging

pyproject.toml is provided for packaging and dependency metadata.

License

MIT