AgentSanchez

## claude-code-gpt-oss-harmony-transformer-litellm-proxy.py
"""
Yes I used AI to make this. This script is taking the response from downstream LLM call, detecting
and replacing harmoney function calls with something claude code can understand. It also outputs some debugging for
File Calls, and token usage. I haven't figured out how to fix the token endpoint in litellm proxy with vertex-ai gpt-oss.
I made this to have fun and compare different models attempting to complete the same kinds of tasks. Use at your own discretion.
I'm sure some of what is used below for harmony conversion can be used in other contexts outside of litellm proxy.

Link to LiteLLM proxy: https://docs.litellm.ai/docs/simple_proxy

oss_tool_fix.py
	"""
	Yes I used AI to make this. This script is taking the response from downstream LLM call, detecting
	and replacing harmoney function calls with something claude code can understand. It also outputs some debugging for
	File Calls, and token usage. I haven't figured out how to fix the token endpoint in litellm proxy with vertex-ai gpt-oss.
	I made this to have fun and compare different models attempting to complete the same kinds of tasks. Use at your own discretion.
	I'm sure some of what is used below for harmony conversion can be used in other contexts outside of litellm proxy.

	Link to LiteLLM proxy: https://docs.litellm.ai/docs/simple_proxy

	oss_tool_fix.py