> I've been using it as an internal tool on many projects
that's interesting, because your LLM decided to put this in the readme:
> Things that didn't work - kept in the repo on purpose.
...which is not what I'd expect to see from a tool that has been battle-tested in a private repo and then exported to a public repo for open-sourcing.
also:
> Caveats
> Single-node Docker on one laptop - no cluster, no replication, no network.
either you're lying about this being an internal tool you've used on real-world clusters/projects, or your LLM is lying on your behalf about this project only having been tested with Docker on a laptop. which is it?
5 comments
[ 4.9 ms ] story [ 20.6 ms ] threadCurious about the premise because ultimately whatever RAM it's using is going to correlate with the data you put in it...
"Show HN" really needs a dedicated subcategory for vibe-coded slop.
that's interesting, because your LLM decided to put this in the readme:
> Things that didn't work - kept in the repo on purpose.
...which is not what I'd expect to see from a tool that has been battle-tested in a private repo and then exported to a public repo for open-sourcing.
also:
> Caveats
> Single-node Docker on one laptop - no cluster, no replication, no network.
either you're lying about this being an internal tool you've used on real-world clusters/projects, or your LLM is lying on your behalf about this project only having been tested with Docker on a laptop. which is it?