livekit-plugins-anthropic: Repeated stream text output after function call

[SleepAt11]

Description: When using livekit-plugins-anthropic with the voice_assistant agent, I've encountered an issue where the LLM repeats the exact same stream text output after triggering a function call.

Steps to reproduce:

1. Set up a project using livekit-plugins-anthropic and voice_assistant agent
2. Implement a scenario that triggers a function call
3. Observe the LLM's output before and after the function call

Expected behavior: The LLM should continue generating new, relevant text after the function call.

Actual behavior: The LLM repeats the exact same stream of text that was output before the function call.

Environment:

• Python version: 3.12.4
• livekit-plugins-anthropic: Installed from latest GitHub code (pip install 'git+https://github.com/livekit/agents.git#subdirectory=livekit-plugins/livekit-plugins-anthropic')
• Other relevant dependency versions: livekit==0.14.0 livekit-agents==0.8.8 livekit-api==0.7.0 livekit_plugins_elevenlabs==0.7.3 livekit-plugins-deepgram==0.6.5 livekit-plugins-openai==0.8.1 livekit-plugins-silero==0.6.4 livekit-protocol==0.6.0

Additional information:

• I noticed that pip version 0.1.1 doesn't work, which is why I'm using the latest GitHub code.

[keepingitneil]

Thanks for the bug report, will look into this today (as well as the deployment issue)