---
title: "AI agent"
type: "glossary-term"
category: "AI & LLMs"
canonical_url: "https://www.graphcanon.com/glossary/ai-and-llms/ai-agent"
updated: "2026-07-09"
---

# AI agent

_Also known as: Agent_

**An AI agent is a system that uses a language model to decide and take actions - calling tools, running steps in a loop - toward a goal, rather than just answering once.**

An **AI agent** wraps a language model in a loop: it observes a goal, decides on an action (often calling an external tool or API), sees the result, and repeats until the task is done. This is the difference between a single answer and autonomous, multi-step work.

Agents rely on tool interfaces to act on the world. Protocols like MCP standardize how an agent discovers and calls those tools.

## In GraphCanon

Agent frameworks are one of the busiest categories in the graph. GraphCanon itself is built to be queried by agents at runtime through its MCP server and JSON API.

## See also

- [AI agents category](/categories/ai-agents)
- [MCP server docs](/docs/mcp)

## Related terms

- [Model Context Protocol (MCP)](/glossary/ai-and-llms/model-context-protocol.md)
- [Generative engine optimization (GEO)](/glossary/discovery-for-agents/generative-engine-optimization.md)

[AI & LLMs](/glossary/ai-and-llms.md) · [All glossary terms](/glossary.md)

---

**Machine-readable endpoints**

- JSON: [`/api/graphcanon`](/api/graphcanon)
- LLM index: [/llms.txt](/llms.txt)
- Full corpus: [/llms-full.txt](/llms-full.txt)

_GraphCanon - The knowledge graph for AI development. https://www.graphcanon.com/_