Agents & knowledge
AI agent
A system that plans and takes actions using models, tools and state.
Also known as: agentic AI
Definition
An AI agent iteratively interprets a goal, chooses actions, uses tools, observes results and continues until it completes or stops a task.
Why it matters
Agent evaluation must assess the full trajectory, permissions and state changes—not only the final text answer.
Related concepts
- System under test
The complete configuration being evaluated—not only the model.
- Evaluation harness
The software and rules used to run and score an eval.
- Tool use
A model selecting and calling external functions or systems.
- Operational fitness
Whether a system works reliably within real service constraints.