# Guardrail Activity

## Overview

The **Guardrail Activity** page provides a unified view of all agentic guardrail detections across your environment. It enables you to monitor guardrail behavior, assess risk levels, and manage detected events efficiently from a single interface.

<figure><img src="/files/PX8qk0xbiOkgh7A2PkpI" alt="" width="563"><figcaption></figcaption></figure>

## Key Metrics

Key metrics summarise guardrail behavior at a glance and help you quickly assess overall security posture and system performance.

### 1. Top Guardrails

Displays the guardrails that are triggered most frequently. This highlights guardrails that are actively engaged and categories where detections are concentrated.

### 2. Guardrail by Severity

Shows detections grouped by severity level, giving clear visibility into the distribution of low, medium, and high-risk events.

### 3. Guardrail (Request vs Response Over Time)

Tracks the number of guardrail messages detected in **requests** versus **responses** over time. This helps distinguish whether detections originate from inbound inputs or outbound agent responses.

### 4. Flagged vs Safe Messages

Shows the trend of messages classified as **flagged** versus **safe** over time, reflecting overall guardrail decision outcomes and stability.

### 5. Guardrail Detection Latency

Displays the **95th percentile detection latency** for guardrail evaluation. It represents the upper bound of guardrail processing time experienced by most requests.

## Guardrail Activity List

The activity list provides a detailed, event-level view of all guardrail detections. Events are grouped into the following categories:

* **Active**
* **Under Review**
* **Ignored**

<details>

<summary>Each activity record contains the following columns:</summary>

| Column             | Description                                                |
| ------------------ | ---------------------------------------------------------- |
| Severity           | Risk level assigned to the detection                       |
| Agentic Component  | Component where the detection occurred                     |
| Host               | Host handling the interaction                              |
| Actor              | Entity initiating the interaction                          |
| Attack Type        | Detected violation or attack category                      |
| Successful Exploit | Indicates whether exploitation was successful (true/false) |
| Collection         | Associated data collection                                 |
| Detected           | Timestamp when the activity was detected                   |

</details>

{% hint style="info" %}
**Search and Filters**

You can refine the activity list using the search bar and the following filters:

* Actor
* URL
* Host
* Type
* Latest attack sub-category
* Successful exploit
  {% endhint %}

## Take Bulk Action

Bulk actions can be performed on multiple guardrail activities at once.

**Steps:**

{% stepper %}
{% step %}
Select activities using the checkbox at the start of each row.
{% endstep %}

{% step %}
Bulk action options appear at the bottom center of the screen.

<figure><img src="/files/TdqvmEuDGgoz9GjaEykC" alt="" width="563"><figcaption></figcaption></figure>
{% endstep %}

{% step %}
Choose one of the following actions:

* Mark selected events for review
* Ignore selected events
* Delete selected events
  {% endstep %}
  {% endstepper %}

## What Next

Navigate to the [Guardrail Activity Detailed View](/agentic-guardrails/how-to/guardrail-activity-detailed-view.md) page to analyze individual detections in greater detail and understand their full context.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://ai-security-docs.akto.io/agentic-guardrails/concepts/guardrail-activity.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.