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

# AI-driven Audits

> Ask the AI to audit a device, integration, or environment, and it builds an interactive audit dashboard right in the chat.

import { CardGroup, Card, Note, Steps, Step, Tip } from '@mintlify/components';

Myrmex doesn't have a static vulnerabilities or CVE screen to scroll through. Instead, you **ask the AI to run an audit** — and the agent assembles a self-contained, interactive audit dashboard right there in the chat, built around the exact question you asked.

## Run an Audit

There are two ways to start:

<CardGroup cols={2}>
  <Card title="Just ask" icon="comments">
    Describe the audit you want in plain language and let Centurion route it to
    the right specialist.
  </Card>

  <Card title="Use the Audit template" icon="clipboard-check">
    Pick the built-in **Audit** prompt template to start from a structured prompt
    instead of a blank one.
  </Card>
</CardGroup>

Scope the audit by `@`-mentioning what it should cover — a device, an integration, or a folder of assets — so the agent audits exactly the right target.

<Steps>
  <Step title="Ask for the audit">
    *"@Centurion audit the security configuration of my domain controller and show me what's misconfigured."*
  </Step>

  <Step title="The agent investigates">
    It runs its audit tools against the target, streaming its reasoning and tool
    calls as it goes.
  </Step>

  <Step title="A dashboard appears">
    The results render as an interactive audit dashboard in the Workspace — a
    self-contained view you can explore, not just a wall of text.
  </Step>
</Steps>

## The Audit Dashboard

The dashboard is generated to fit what the agent found: summaries, findings, and detail you can drill into, all rendered inline. Because a generated audit is saved as an **Audit** artifact, you can reopen it later from the **Artifacts** section of the Directory.

<Note>
  Audits are point-in-time: each one answers the question you asked, when you
  asked it. For an always-on measure of how your environment is doing, use
  Security Posture instead.
</Note>

## Audits vs. Continuous Posture

An audit is something you run on demand for a specific question. **Security Posture** is the continuous counterpart — a standing score and grade for your environment, refreshed over time and mapped to compliance frameworks.

<CardGroup cols={2}>
  <Card title="Security Posture" icon="shield-halved" href="/documentation/security-posture/overview">
    Continuous scoring, a priority queue, and a compliance heatmap for your whole
    environment.
  </Card>

  <Card title="Chatting with the AI" icon="comments" href="/documentation/ai-workspace/chat">
    Prompt templates, `@`-mentions, and how conversations work.
  </Card>
</CardGroup>