# Classifiers

> Use deterministic command-safety review as bounded evidence inside Harness gate policy.

- Path: `Modules > Classifiers - Permission Classification`
- Human: https://looprig.com/docs/modules/classifiers
- Machine index: https://looprig.com/llms.txt

## Repository

| Field | Value |
| --- | --- |
| Repository | `github.com/looprig/classifiers` |
| Version | `v0.2.0` |
| GitHub | [looprig/classifiers](https://github.com/looprig/classifiers) |

## Description

Use deterministic command-safety review as bounded evidence inside Harness gate policy.

## Where it fits

Classifiers are useful on their own when an application needs deterministic command-safety evidence. Within Looprig, they support permission decisions around [Harness](/docs/modules/harness.md) and prepared [Tools](/docs/modules/tools.md). A classifier reports bounded evidence; it does not grant authority or replace the gate that makes the final decision.

## Dependencies

- [Core](/docs/modules/core.md)
- [Harness](/docs/modules/harness.md)
- [Inference](/docs/modules/inference.md)

## Dependents

None.