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

Archive

• Create zip from files and directories

• Extract zip to directory (with ZipSlip protection)

Extract Text

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

Related MCP server: PDF2MD MCP Server

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.