The evaluation table currently shows "too many requests" when an LLM provider returns a quota error. This message confuses users because it doesn't explain what went wrong or how to fix it.
Expected behavior
The evaluation table should display the actual error from the provider (like "OpenAI rate limit exceeded"). Users can hover over the error to see technical details such as the HTTP status code and full error response.
Why this matters
Error messages should be clear and actionable. When users see "too many requests," they don't know if the problem is with Agenta's infrastructure, their API key, or their provider's quota limits. Showing the provider's error message immediately tells them what's wrong and what they need to do about it.
The evaluation table currently shows "too many requests" when an LLM provider returns a quota error. This message confuses users because it doesn't explain what went wrong or how to fix it.
Expected behavior
The evaluation table should display the actual error from the provider (like "OpenAI rate limit exceeded"). Users can hover over the error to see technical details such as the HTTP status code and full error response.
Why this matters
Error messages should be clear and actionable. When users see "too many requests," they don't know if the problem is with Agenta's infrastructure, their API key, or their provider's quota limits. Showing the provider's error message immediately tells them what's wrong and what they need to do about it.