> ## Documentation Index
> Fetch the complete documentation index at: https://growthx-chore-remove-output-wrapper.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Score LLM output, control workflow flow, and test quality across datasets

Evaluators score content — usually LLM output — and return results with a value, a confidence score, and optional reasoning. Output uses evaluators in two distinct ways:

**[Evaluator Step](/evaluators/evaluator-step)** — An evaluator that runs *inside* your workflow as a step. Your workflow generates something, evaluates it, and decides what to do next: retry if quality is low, skip if confidence is high, or branch to a different path. This generate-evaluate-retry loop is how you build self-correcting workflows.

```typescript workflow.ts theme={null}
const summary = await generateSummary( company );
const quality = await judgeSummaryQuality( { summary, companyName: company.name } );

if ( quality.value === true && quality.confidence >= 0.7 ) {
  return summary;
}
// retry or take a different path...
```

**[Evaluation Workflow](/evaluators/workflow-evaluators)** — A separate workflow that tests another workflow's quality across a dataset of test cases. You define evaluators with `verify()`, wire them into an eval workflow, and run them from the CLI. Use this for regression testing, CI/CD quality gates, and systematic quality monitoring — without modifying your production workflow code.

```bash theme={null}
npx output workflow test my_workflow --dataset golden_set
```

## Which one do I need?

| I want to...                                      | Use                                                                                      |
| ------------------------------------------------- | ---------------------------------------------------------------------------------------- |
| Have my workflow check and improve its own output | [Evaluator Step](/evaluators/evaluator-step)                                             |
| Test my workflow against a set of known inputs    | [Evaluation Workflow](/evaluators/workflow-evaluators)                                   |
| Add quality gates that retry on failure           | [Evaluator Step](/evaluators/evaluator-step)                                             |
| Run evals in CI/CD before deploying               | [Evaluation Workflow](/evaluators/workflow-evaluators)                                   |
| Use both in the same project                      | Start with evaluator steps in your workflow, then add an evaluation workflow for testing |

Both approaches use evaluators under the hood — the difference is *where* they run and *what* they control.

## What's Next

<CardGroup cols={2}>
  <Card title="Evaluator Step" icon="layer-group" href="/evaluators/evaluator-step">
    **Evaluator Step.** Build evaluators and use them inside your workflows for self-correction.
  </Card>

  <Card title="Evaluation Workflow" icon="flask-vial" href="/evaluators/workflow-evaluators">
    **Evaluation Workflow.** Test workflow quality across datasets from the CLI.
  </Card>

  <Card title="LLM-as-a-Judge Best Practices" icon="lightbulb" href="/evaluators/best-practices">
    **LLM-as-a-Judge Best Practices.** Writing effective judge prompts, grading scales, and common pitfalls.
  </Card>
</CardGroup>