* Add eslint errors
- @typescript-eslint/strict-boolean-expressions
- eqeqeq
* Remove boxen
* Remove more boxing (bg) in texts and make error messages more consistent
* Make get interactive as autocomplete instead of list
* Misc log cleanup
* Fix rebasing
* Slight tweak for lms load
---------
Co-authored-by: Will Burford <will@lmstudio.ai>
* Refactor start and stop server to use a provider
* Fix circular dependencies
* Actually correct arg
* Bring back port arg
* Edit stop command too
* Addressed comments
* Addressed comments
* Addressed comments
* Environment support for multiple hosts or ports
* Better error handling
* Minor changes
* Correct handling for local
* Add inspect command
* Address comments
* Address comments more
* Open auth URL automatically in browser on login
* Pin older version of open for CJS support
* Vendor openUrl
* Remove open dep
* Wrap in try/catch
* Promisify
* feat(cli): add --verbose flag to chat command
* Address review comments: add types, rename --verbose to --stats, remove emoji and GPU layers stat
- Add proper types for displayVerboseStats function parameters
- Rename --verbose flag to --stats to avoid conflict with global verbose flag
- Add short flag -t for --stats
- Remove emoji from stats output for cleaner appearance
- Hide numGpuLayers stat as it's not currently populated
* Update src/subcommands/chat.ts
Co-authored-by: Yagil Burowski <yagil@elementlabs.ai>
* Remove short stats flag
* Update Prediction Stats title
* Add second new line ln26
* Remove short stats flag again
---------
Co-authored-by: Yagil Burowski <yagil@elementlabs.ai>
* Allow lms dev to not notify
* Add flags to make private artifacts / write back revision number
* lms push --private doc
* Improve lms dev --no-notify message
* first working chat example "lms ask sup" or "lms chat"
* print streaming response instead of waiting until the end
* visual polish
* removed ask alias, chat only. --prompt or -p to provide initial prompt (quits after response). model selection i.e. "lms chat phi-4"
* use existing Chat class for context management, use existing optionalPositional() for model arg
* prettier changes, added types, added quit instructions
* allow combining -p with stdin for processing piped content with a prompt