What is a Tool in the context of AI agents?

• A. A Set of Internal Thoughts Used for Planning
• B. A Static Dataset Used to Train the Model
• C. A Function Given to the LLM That Performs a Specific Task
• D. A Feedback Loop Mechanism for Error Handling

Answer: (C)

Tools in AI agents are external functions the language model can call to perform specific tasks. They extend what the agent can do beyond pure text generation by interfacing with calculators, search services, databases, file systems, or APIs. The agent might plan to invoke a tool with certain inputs, receive the tool’s output, and then use that result to continue reasoning or produce a final answer. This distinguishes tools from training data, which is static information the model learned during training and cannot be invoked at runtime; and from internal thoughts or planning that occur inside the model without calling an external function. It’s also not a feedback loop mechanism, which is about adjusting behavior rather than performing a concrete operation. A Tool is simply a function provided to the LLM that performs a specific task, enabling real-world actions and computations.