Function Calling & Tools

This section provides background on Function Calling & Tools (opens in a new tab) using the MuleSoft AI Chain (MAC) Project.

The concept of Tools, also known as Function Calling allows the LLM to, when necessary, one or more available tools, usually defined by the developer. Natively the LLM provides a tool execution request of tools to be used to answer a users query.

In the MAC Project, we define tools as API resources running on the MuleSoft Anypoint Runtimes, and published on Anypoint Exchange. There are 2 main operations available:

• Tools Use AI Service: The Tools | Use AI Service operation is useful if you want to create autonomous agents that can use external tools whenever a prompt cannot be answered directly by the AI model. A prerequisite for using Tools is to prepare a tools.config.json file, which includes all the necessary API information for the operation to execute APIs successfully.
• Tools Use AI Native: The Tools | Use Native Template operation is useful if you want to create autonomous agents that can use external tools whenever a prompt cannot be answered directly by the AI model. This operation only provides a request to execute the Tools provided in the payload. It doesn't execute them.

Tools Use AI Service

This operation has implemented tool execution out-of-the-box. You can provide the available tools in a tool.config.json file, which the LLM will assess and executes based on the users query and provided tool implementation details.

Supported by Connectors:

• MuleSoft AI Chain
• Einstein AI
• Amazon Bedrock

Tools Use AI Native

This operation has implemented tool execution out-of-the-box. You can provide the available tools in a tool.config.json file, which the LLM will assess and executes based on the users query and provided tool implementation details.

Supported by Connectors:

• MuleSoft AI Chain
• MAC Inference