sprine

Thank you. I'm going to start using this too. I prefer reading offline as well.

I've been working on a Ontario open data MCP: https://github.com/sprine/ontario-data-mcp

There is a wealth of data that's behind CSVs and other data formats. This uses DuckDB as a common (local) database to cache and run queries against, and enables going across datasets for insights using LLMs.

I wish something existed in this space. I used Coati Software's Sourcetrail for a couple of years. Unfortunately it was discontinued. It was a wonderful piece of software that indexed a code repository, and exposed an interface to explore it interactively. At least for me, it significantly improved the understanding and legibility of code.

The code is in an archived state (https://github.com/CoatiSoftware/Sourcetrail). Searching for the software on Google shows some screenshots.