# Guardrail actions

Guardrail actions are what a guardrail does when a policy matches. The common ones are block, rewrite, redact, hold for approval, terminate the session, flag and alert, and each policy gets its own.

## The actions

| Action | What it does | When it fits |
| --- | --- | --- |
| Block | Stops the message, reply or tool call | Clear and serious violations |
| Rewrite | Replaces the reply with one that complies | A reply that is useful once fixed |
| Redact | Masks a span and passes the rest | Personal data and secrets |
| Hold for approval | Pauses a tool call until a person confirms | Agent actions that cannot be undone |
| Terminate | Ends the session | An attack in progress |
| Flag | Passes it and marks it for review | New policies and borderline cases |
| Alert | Notifies a person or a system | Events someone must see now |

## How to choose

Match the action to the cost of a mistake. A new [policy](/glossary/guardrail-policy) can start on flag and move to block once its [false positives](/glossary/guardrail-false-positives) are low. A tool call has to be stopped or held before it runs. Holding it puts a [human in the loop](/glossary/human-in-the-loop).

## With White Circle

A policy in [Control](/products/control) can block, rewrite, redact, hold a tool call for confirmation, terminate the session, flag, or send an alert.

