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

# Guardrails

> One switch that screens agent replies for harmful, biased, or inappropriate content before they reach a customer.

## What is this?

**Guardrails** are safety checks that run on the agent's replies — like a seatbelt, not something you tune, just something you keep on.

## What can I do here?

* Toggle **Enable Guardrails** on or off.

## How to use it

Turn it on and leave it on. There's no real reason to disable guardrails for an agent talking to real customers — doing so removes a safety net without adding any capability. If a guardrail is blocking a reply you believe is legitimate, fix the underlying [Instructions](https://wabis.mintlify.site/workspace/ai-agents/tabs/instructions.md) rather than turning the whole check off.

> [!WARNING]
>
> Disabling guardrails removes the platform's safety checks on this agent's replies entirely. Use with caution in production — this protects your brand reputation and keeps customer interactions safe and compliant.

## Related docs

* [Instructions](https://wabis.mintlify.site/workspace/ai-agents/tabs/instructions.md)
* [Test](https://wabis.mintlify.site/workspace/ai-agents/tabs/test.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.