LinuxCommandLibrary
GitHubF-DroidGoogle Play Store

marky

Convert documents of many formats into Markdown

TLDR

Convert a PDF and print the Markdown to stdout
$ marky [document.pdf]
copy
Convert a Word document and save to a file
$ marky [document.docx] --output [converted.md]
copy
Convert an Excel spreadsheet using the short form of the output flag
$ marky [data.xlsx] -o [table.md]
copy
Convert a PowerPoint presentation
$ marky [presentation.pptx] -o [slides.md]
copy
Convert an HTML page to Markdown
$ marky [webpage.html] -o [content.md]
copy
Convert a Jupyter notebook
$ marky [notebook.ipynb] -o [notebook.md]
copy

SYNOPSIS

marky INPUT [-o OUTPUT]

DESCRIPTION

marky is a small Go command-line tool (and embeddable library) that converts documents into clean, readable Markdown. It understands CSV, EPUB, HTML, Jupyter notebooks (.ipynb), Microsoft Word (.docx), Microsoft Excel (.xlsx), PDF, and PowerPoint (.pptx) files.The input format is detected from the file extension, so a single command — marky somefile.ext — is usually enough. Without -o, Markdown goes to stdout so marky can be chained with other tools (marky report.pdf | pandoc -o report.html). The project also ships a marky-mcp companion that exposes the same conversion functions over the Model Context Protocol for use with coding agents.

PARAMETERS

-o FILE, --output FILE

Write the generated Markdown to FILE. Without this flag the result is printed to standard output.
-h, --help
Show usage information.
-v, --version
Print the version and exit.

INSTALLATION

marky is distributed as a Go program:

$ go install github.com/flaviodelgrosso/marky/cmd/marky@latest
copy
A Go 1.24 or newer toolchain is required. The companion MCP server installs with:
$ go install github.com/flaviodelgrosso/marky/marky-mcp@latest
copy

CAVEATS

Conversion fidelity depends on the underlying libraries (html-to-markdown, excelize, PDF parsers); complex layouts, equations, and embedded images may not round-trip perfectly. Very large Excel workbooks can consume significant memory because marky loads the whole document before conversion.

HISTORY

marky is authored by Flavio Del Grosso and released under the ISC license. It combines several existing Go libraries into one cohesive CLI and MCP server to make feeding documents to AI agents straightforward.

SEE ALSO

pandoc(1), html2md(1), pdftotext(1), docx2txt(1)

Copied to clipboard
Kai