- Add tests to verify warnings for unknown models in OpenAI, Anthropic, and Google clients - Confirm no warnings are issued for known models in OpenAI client - Verify OpenRouter provider allows custom models without warnings - Implement helper assertion for checking single unknown model warning - Ensure tests patch underlying LLM constructors for isolation and inspection |
||
|---|---|---|
| .. | ||
| agents | ||
| dataflows | ||
| graph | ||
| llm_clients | ||
| __init__.py | ||
| default_config.py | ||