Beyond a chatbot
A chatbot only replies. An agent can call tools: look up an order, list low-stock products, draft an email. The model decides which tool to use; the tools do the actual work.
Why the tools matter
An agent is only as safe as its tools. If every tool is narrow, checked and logged, the agent cannot do much harm even when the model is wrong.
Part of: Understanding AI agents
Next recommendedTools, permissions and guardrailsRelated
Last updated 2026-09-25. Something unclear or wrong? Tell us.