But SQLite is notoriously 35% faster than the filesystem [0], so why not use that?
My biggest success is a Roslyn method that takes a .NET solution and converts it into a SQLite database with Files, Lines, Symbols, and References tables. I've found this approach to perform substantially better than a flat, file-based setup (i.e., like what Copilot provides in Visual Studio). Especially, for very large projects. 100+ megs of source is no problem. The relational model enables some really elegant [non]queries that would otherwise require bespoke reflection tooling or a lot more tokens consumed.
In practice this allows for me to combine multiple, complex data sources with a constant number of tools. I can add a whole new database and not add a new tool. My prompts are effectively empty aside from metadata around the handful of tools it has access to.
This only seems to perform well with powerful models right now. I've only seen it work with GPT5.x. But, when it does work it works at least as well as a human given access to the exact same tools. The bootstrapping behavior is extremely compelling. The way the LLM probes system tables, etc.
The tasks this provides the most uplift for are the hardest ones. Being able to make targeted queries over tables like references and symbols dramatically reduces the number of tokens we need to handle throughout. Fewer tokens means fewer opportunities for error.
Is the Roslyn method called as part of the build/publish?
AgentFS https://agentfs.ai/ https://github.com/tursodatabase/agentfs
Which sounds like a great idea, except that is uses NFS instead of FUSE (note that macFUSE now has a FSKit backend so FUSE seems like the best solution for both Mac and Linux).