lmstudio-ai/lms — reverse-engineered prompt
Reverse engineered prompt
Build me a simple command line tool for LM Studio called lms. I want it to feel like the terminal helper for managing my local LM Studio setup, not a complicated developer project.
It should let me check whether LM Studio is running, start and stop the local API server, see which models are downloaded, see which ones are currently loaded for inference, load a model, unload one model or unload everything, create a new starter project, and stream logs while LM Studio is running. Make the command help really clear so I can discover things by typing help, and support JSON output for the model listing commands where that already makes sense.
I also want the usual install to path experience so the command is easy to run from a fresh terminal. If this repo only works properly as part of the larger LM Studio JS monorepo, please wire it up the right way and use that setup instead of forcing a standalone build. Look up current LM Studio docs online if you need to.
Want more depth? Deep Reverse