Function call fail when using Claude LLM with v0.11.1

[zamia]

when using claude llm, the function call feature fails with newest version 0.11.1.

Error message:

livekit.agents._exceptions.APIStatusError: Error code: 400 - {'type': 'error', 'error': {'type': 'invalid_request_error', 'message': 'Requests which include `tool_use` or `tool_result` blocks must define tools.'}} {"pid": 20121, "job_id": "AJ_XfSd2bnigEG9"}

Reproduce:

1. use the weather example, change to claude llm and upgrade to the newest version.
2. test the conversation.

Seems the second round to call to LLM fails because of losing tools context.

[zamia]

If anyone met same problem, you could work around by setting the max_nested_fnc_calls to a bigger value than 1 when initializing VoicePipelineAgent.