> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Criteria

> Evaluate agents against custom criteria

Criteria evaluation measures agent output against custom standards.

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
    subgraph "Criteria Evaluation"
        O[📤 Output] --> E[📊 Evaluate]
        E --> C1[✓ Clarity]
        E --> C2[✓ Accuracy]
        E --> C3[✓ Helpfulness]
    end
    
    classDef output fill:#6366F1,stroke:#7C90A0,color:#fff
    classDef criteria fill:#10B981,stroke:#7C90A0,color:#fff
    
    class O output
    class E,C1,C2,C3 criteria
```

## Quick Start

<Steps>
  <Step title="Define Criteria">
    ```rust theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    use praisonai::CriteriaEvaluator;

    let evaluator = CriteriaEvaluator::new()
        .criterion("clarity", "Is the response clear and easy to understand?")
        .criterion("accuracy", "Is the information factually correct?")
        .criterion("completeness", "Does it fully answer the question?")
        .build();

    let scores = evaluator.evaluate(&response);
    for (name, score) in scores {
        println!("{}: {:.0}%", name, score * 100.0);
    }
    ```
  </Step>
</Steps>

***

## Common Criteria

| Criterion    | Description            |
| ------------ | ---------------------- |
| Clarity      | Easy to understand     |
| Accuracy     | Factually correct      |
| Completeness | Fully addresses query  |
| Conciseness  | No unnecessary content |
| Helpfulness  | Actionable and useful  |

***

## Related

<CardGroup cols={2}>
  <Card title="Evaluation" icon="chart-bar" href="/docs/rust/evaluation">
    Evaluation system
  </Card>

  <Card title="Optimizer" icon="wand-sparkles" href="/docs/rust/optimizer">
    Auto-improvement
  </Card>
</CardGroup>