This node creates an agent based on (Azure) OpenAI chat models that support function calling (e.g. the 0613 models) and can be primed with a custom system message.
The system message plays an essential role in defining the behavior of the agent and how it interacts with users and tools. Before adjusting other model settings, it is recommended to experiment with the system message first, as it has the most significant impact on the behavior of the agent.
An agent is an LLM that is configured to pick a tool from a set of tools to best answer the user prompts, when appropriate.
For Azure : Make sure to use the correct API, since function calling is only available since API version '2023-07-01-preview'. For more information, check the Microsoft Documentation
Note : These agents do not support tools with whitespaces in their names.
Note : If you use the Credentials Configuration node and do not select the "Save password in configuration (weakly encrypted)" option for passing the API key for the chat model, the Credentials Configuration node will need to be reconfigured upon reopening the workflow, as the credentials flow variable was not saved and will therefore not be available to downstream nodes.