Tools for AI Agentss
Tools are the functions an agent can call to act on the world, such as searching, querying a database or sending a request. Good tool design means clear names, typed parameters and useful errors. The Model Context Protocol, or MCP, standardises how tools are exposed to agents.
Stage: Foundations · Part of the AI Agents roadmap · Last reviewed 19 November 2026
What to learn in Tools
- Write a tool schema with a precise name, description and typed parameters
- Return errors from a tool in a form the model can recover from
- Expose an existing function to an agent through an MCP server
- Limit what each tool can read or change using the narrowest permissions
- Decide how many tools to give an agent before choices degrade
Tools interview questions
What is the Model Context Protocol and why does it matter?
MCP is an open protocol that lets an AI application discover and call tools, resources and prompts from external servers in a standard way. It matters because one MCP server can work with many agent clients, so you write the integration once instead of per framework.
Why do tool descriptions affect agent quality so much?
The model picks tools only from their names, descriptions and parameter schemas. Vague or overlapping descriptions cause wrong calls or skipped calls. Specific descriptions that state when to use a tool, and when not to, usually improve accuracy more than prompt tweaks do.
Where Tools fits in the AI Agents roadmap
Tools belongs to the Foundations stage (intermediate level), which takes about 23 hours. What an agent is.
Other units in this stage
Back to the full AI Agents roadmap