catch litellm no deployment error (token ratelimit / ValueError) and raise LLMProviderErrorRetryableTask

[wintonzheng]

🚀	This description was created by Ellipsis for commit 20241d5

Summary:

Handles ValueError in llm_api_handler_with_router_and_fallback by logging and raising LLMProviderErrorRetryableTask.

Key points:

• Added ValueError exception handling in llm_api_handler_with_router_and_fallback.
• Logs "LLM token limit exceeded" when ValueError occurs.
• Raises LLMProviderErrorRetryableTask for ValueError.
• Change made in skyvern/forge/sdk/api/llm/api_handler_factory.py.

---

Generated with ❤️ by ellipsis.dev

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to 20241d5 in 15 seconds

More details

• Looked at 19 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. skyvern/forge/sdk/api/llm/api_handler_factory.py:113

• Draft comment:
  The log message "LLM token limit exceeded" is misleading as ValueError can be raised for other reasons. Consider making the log message more generic or checking the specific condition that raises ValueError.
• Reason this comment was not posted:
  Decided after close inspection that this draft comment was likely wrong and/or not actionable:
  The comment is relevant to the change made in the diff, as it addresses the new exception handling block for ValueError. The concern about the log message being potentially misleading is valid because ValueError can indeed be raised for various reasons, not just token limit issues. The comment suggests a specific improvement, which aligns with the rules for actionable and clear code quality refactors.
  The comment might be considered speculative if there is a specific context in which ValueError is only raised for token limit issues. However, without such context, the comment seems valid.
  The comment is valid unless there is specific documentation or code that guarantees ValueError is only raised for token limit issues in this context.
  The comment should be kept as it addresses a potential issue with the new code change and suggests a clear improvement.

Workflow ID: wflow_lHRslaHrvsP6ImTQ

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.