dcdpr / jp / 15595635693

58%

push

github

feat(cli): Improved model detection during workspace initialization (#145)

The `jp init` command now automatically detects and configures a default
model for the workspace instead of using a hard-coded default. The
detection logic prioritizes models in the following order:

1. Environment variable `JP_LLM_MODEL_ID`,
2. locally available Ollama models (preferring `llama`, `gemma`, then
`qwen` variants),
3. and finally cloud providers based on available API keys (`Anthropic`,
`OpenAI`, `Gemini`).

When a model is detected, users see a confirmation message with a note
about how to change it later. This eliminates the need for users to
manually configure a model before starting their first conversation.

Signed-off-by: Jean Mertz <git@jeanmertz.com>

1 of 51 new or added lines in 2 files covered. (1.96%)

3586 of 9278 relevant lines covered (38.65%)

4.69 hits per line