In the core agent, we need better error handling of the multiple possible problems and errors we can get from the LLM provider and liteLLM in general.
For example,
- If the authentication fails, catch the error
- If the provider uses another different API that is badly supported by liteLLM
- If the model response by content policy protection limiting the use of shell and other tool to basically do the work
In the core agent, we need better error handling of the multiple possible problems and errors we can get from the LLM provider and liteLLM in general.
For example,