Ollama integration

[swerner]

Add Ollama as a model provider. Ollama provides OpenAI compatible unified API access to hundreds of open source models. The latest models like Gemma4 are getting incredibly good and also run well on consumer hardware. Similar to #70 - would give users access to a much wider selection of models and provide a free token usage, especially for less critical tasks.

[brynary]

Hi @swerner -- Model providers are now fully configurable via settings. I added an Ollama example here for illustrative purposes: https://github.com/fabro-sh/fabro/blob/main/lib/crates/fabro-model/src/catalog/providers/ollama.toml

Settings are layered so you should be able to enable Ollama from your server's settings.toml file like this:

[providers.ollama]
enabled = true

After that, you'll just want to add whatever Ollama models you want to expose to Fabro as extra blocks. (Model auto-discovery is captured in #267)