📦

File Converter MCP

A Python-based MCP (Model Context Protocol) server powered by FastMCP that leverages Pandoc for flexible document conversion between various formats (Markdown, DOCX, PDF, HTML, etc.), designed for use with AI agents like Claude. Includes Docker configuration for easy deployment.

0 installs

23 stars

15 forks

Trust: 54 — Fair

Devtools

Installation

npx file-converter-mcp

Ask AI about File Converter MCP

Powered by Claude · Grounded in docs

I know everything about File Converter MCP. Ask me about installation, configuration, usage, or troubleshooting.

0/500

Documentation

File Ops

简体中文 | English

A local file operations skill for AI agents. Convert, inspect, archive, and extract text from files on the local machine using a bundled Python script with structured JSON output.

What It Does

Convert

Image formats: png, jpg, webp, gif, bmp, tiff
pdf -> docx
xlsx -> csv
xls -> csv (requires xlrd, not installed by default)
docx -> pdf (requires LibreOffice)
html/md/markdown -> pdf (requires wkhtmltopdf)

Inspect

Basic metadata: size, MIME type, permissions, last modified
Image: dimensions, color mode, format
PDF: page count
Excel: sheet names, row counts

Extract Text

PDF, DOCX, XLSX, HTML, Markdown, image EXIF metadata

Quick Setup

bash skills/file-ops/scripts/setup.sh

Usage

Health check:

skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py health

Convert:

skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py convert --input /path/input.pdf --to docx

Inspect:

skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py inspect --input /path/file.pdf

Archive:

skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py archive --input /path/dir --output /path/out.zip
skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py archive --input /path/archive.zip --output /path/dest --extract

Extract text:

skills/file-ops/.venv/bin/python skills/file-ops/scripts/file_ops.py extract-text --input /path/document.pdf

Use --output /path/output.ext with convert if you do not want the output written next to the input file.