# Tool Definition

> A note that tells a model how to call a tool. It lists the inputs.

**Field:** Artificial Intelligence  
**Category:** AI Agents  
**Also known as:** Tool Schema, Function Schema, Function Definition (LLM)

## Think of it like

It is like the label on a light switch. It says what the switch does and when to use it.

## Definition

A tool definition is the short note that tells a model a tool exists and how to call it. It has a name, a plain description of what the tool does, and a schema for its inputs. The schema lists each input, its type and which ones are required. The model never sees the tool's code, only this note. So a clear description is the main thing that decides whether the model picks the right tool.

## Example

Emma adds a weather tool to her app. Its note says the tool needs a city. It may also take a unit, celsius or fahrenheit.

## Example code

```json
{
  "name": "get_weather",
  "description": "Get the current weather in a city.",
  "input_schema": {
    "type": "object",
    "properties": {
      "city": { "type": "string" }
    },
    "required": ["city"]
  }
}
```

One tool, one required input. The model reads only these lines, never the code behind them.

## Why it matters

The model picks tools by their description. A good one gets the right tool called with the right inputs.

## Common mistake

Writing a one-line description. The model then guesses when to use the tool and what each input means.

## Three principles

- **Name it.** A short, clear name with no spaces.
- **Describe it.** What it does and when to use it.
- **List the inputs.** Each input, its type, and if it is required.

## Related terms

- [Tool Use (Function Calling)](https://learn.benibix.com/term/ai/tool-use)
- [Tool Result](https://learn.benibix.com/term/ai/tool-result)
- [Structured Output (LLM)](https://learn.benibix.com/term/ai/structured-output)
- [JSON Schema](https://learn.benibix.com/term/development/json-schema)
- [Agent Skill](https://learn.benibix.com/term/ai/agent-skill)

![A request carries two tool definitions to the model. The model sends back two tool calls at once, each with an id. Your code runs both and returns two tool results with the same ids. Below, a separate lane shows code the model writes running in a locked sandbox.](https://learn.benibix.com/diagrams/ai7-tool-calls.svg)

---

Source: https://learn.benibix.com/term/ai/tool-definition
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-19
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
