You could have your Coding agent use this to do it for you! Really though it's about using the right tool for the job, and this seems like a better choice to use a tool like this so you don't have your agent burn tons of tokens imperfectly working through all your documents at a slower pace.
Personally I'll be giving this a whirl as part of my note distillation process. I end up with hundreds of pages of PDFs and docx files that I'm sure would be easy to convert with a dedicated tool.
My first thought as well! Even glancing through their website I don’t see an FAQ section or even a section just answering what value addition it brings over pandoc?
Related: https://github.com/microsoft/markitdown
Why do I need this when I can use a coding agent to do it for me?
You could have your Coding agent use this to do it for you! Really though it's about using the right tool for the job, and this seems like a better choice to use a tool like this so you don't have your agent burn tons of tokens imperfectly working through all your documents at a slower pace.
Personally I'll be giving this a whirl as part of my note distillation process. I end up with hundreds of pages of PDFs and docx files that I'm sure would be easy to convert with a dedicated tool.
Why would you make a non-deterministic agent do a task like this if an exact deterministic command line tool exists for it?
so I don't bloat my system with random software that require maintenance and are a potential attack vector?
why would you to waste tokens when you already have open source options...
Pandoc?
My first thought as well! Even glancing through their website I don’t see an FAQ section or even a section just answering what value addition it brings over pandoc?