LLM setup

eunha supports three LLM backends: OpenAI, Anthropic, and Ollama (local).

All keys are stored in ~/.eunha/config.toml (permissions 0600). They are never written to the database.

OpenAI

[llm]
provider = "openai"
api_key = "sk-..."
model = "gpt-4o-mini"   # optional, default: gpt-4o-mini

Anthropic

[llm]
provider = "anthropic"
api_key = "sk-ant-..."
model = "claude-haiku-4-5-20251001"   # optional

Ollama (local)

Install Ollama and pull a model with JSON mode support (e.g. llama3.2):

ollama pull llama3.2

[llm]
provider = "ollama"
model = "llama3.2"
base_url = "http://localhost:11434"   # optional, this is the default

Caution

Ollama is best-effort. If the model doesn’t support JSON mode, eunha will fail with a clear error rather than producing garbage descriptions.

Describing repos

Once configured, navigate to any repo in the library and press d to describe it. Press shift-D to re-describe (regenerate). Press shift-A to batch-describe all undescribed repos.