Lesson 3 of 7
Giving Agents Tools
Tools turn talk into action. Learn how an agent chooses a tool, fills in its inputs and uses the result.
Learn it
A tool is a function with a name, a description and inputs — like getWeather(city).
The model reads the descriptions and decides which tool fits the task.
Your code actually runs the tool and hands the result back to the model.
Key terms
- Tool schema
- The name, description and parameter types describing a tool.
- Tool call
- The model's request to run a tool with specific arguments.
- Least privilege
- Give an agent only the access it truly needs.
- Human-in-the-loop
- A person approves risky actions before they happen.
Trace a tool call
- 1User: 'Will I need an umbrella in Prague tomorrow?'
- 2Model picks: getWeather(city='Prague', day='tomorrow')
- 3App runs: Returns {rain_chance: 80%}
- 4Model answers: 'Yes — 80% chance of rain. Take an umbrella!'
Defining a tool (JavaScript)
javascriptconst tools = {
getWeather: {
description: "Get tomorrow's weather for a city",
parameters: { city: "string" },
run: async ({ city }) => {
return { city, rainChance: 0.8 }; // pretend API
},
},
};The description is what the model reads. The run function is what your app executes.
Try it
Match each task to the best tool
Challenge
Design 3 tools for your agent. For each: name, one-sentence description, parameters, and whether it needs human approval. Explain why.
Pick whichever way suits you — every mode earns the same bonus XP.
Write at least 40 more characters to submit.
Mark your own work
Guided walkthrough — 0/2 clues revealed
- Clue 1 locked — reveal it only if you get stuck.
- Clue 2 locked — reveal it only if you get stuck.
Each clue costs 4 XP (never below 17 XP). You'd earn 33 XP right now.