This PR adds support for pre-execute prompts when registering functions for OpenAILLM. These prompts are sent to the LLM before executing a function call, allowing for additional context or instructions to be provided to the user before the function executes.
For instance, we can instruct the LLM to inform the user that "we're getting the information, please hold on" before executing RAG.
This PR adds support for pre-execute prompts when registering functions for OpenAILLM. These prompts are sent to the LLM before executing a function call, allowing for additional context or instructions to be provided to the user before the function executes.
For instance, we can instruct the LLM to inform the user that "we're getting the information, please hold on" before executing RAG.