> ## 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.

# Remediation & risk acceptance

> The actions you take on a Security Posture finding — analyze it, remediate it through Perseus or Hydra, accept the risk behind a signed document, or reactivate a silenced finding later.

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

Security Posture shows you *what* is wrong; this is how you act on it. Every finding in the [priority queue and findings table](/documentation/security-posture/overview#what-the-dashboard-shows) carries a set of actions, and each one runs through chat so you can watch [Athena](/documentation/multi-agent-system/athena) and the specialist agents do the work.

## The actions on a finding

<CardGroup cols={2}>
  <Card title="Analyze" icon="magnifying-glass">
    Open a deeper analysis of the finding in chat, without changing anything.
  </Card>

  <Card title="Remediate" icon="wrench">
    Send the finding to chat, where an agent fixes the underlying problem.
  </Card>

  <Card title="Silence / accept risk" icon="file-signature">
    Formally accept a finding you've decided to live with, behind a signed
    document.
  </Card>

  <Card title="Reactivate" icon="rotate-left">
    Bring a silenced finding back into scope when circumstances change.
  </Card>
</CardGroup>

## Analyze

**Analyze** opens the finding in chat and asks Athena to explain it — what the issue is, why it matters, which targets it affects, and how it maps to your frameworks. Nothing changes on your hosts or integrations; this is a read-only conversation you use to decide what to do next.

## Remediate

**Remediate** sends the finding — and the affected target — into [chat](/documentation/ai-workspace/chat) as `@`-mention chips and hands the fix to the right specialist:

<CardGroup cols={2}>
  <Card title="On a host — Perseus" icon="server" href="/documentation/multi-agent-system/agent-endpoint-specialist">
    Perseus applies the fix on the device through the Myrmex agent.
  </Card>

  <Card title="On an integration — Hydra" icon="plug" href="/documentation/multi-agent-system/integration-specialist-agent">
    Hydra applies the fix on the connected tool.
  </Card>
</CardGroup>

As with every operational action, the agent proposes the fix and carries it out in front of you, so you stay in control of what actually runs.

## Accept a risk

Not every finding gets fixed. When you choose to live with one, **silence** it — but Athena gates the decision so an accepted risk is always a documented, signed one.

<Steps>
  <Step title="Athena drafts the document">
    Athena generates a **risk-acceptance document** that records the finding,
    the target, and the justification.
  </Step>

  <Step title="Sign it">
    The document has to be **signed** before the finding is silenced.
  </Step>

  <Step title="A critical finding needs a second approver">
    If the finding is **critical**, a second approver must sign off before it
    can be silenced.
  </Step>
</Steps>

<Note>
  This is the "AI proposes, you approve" pattern applied to risk: Athena
  prepares the decision, but a human signs before anything is silenced.
</Note>

Once silenced, the finding drops out of your active priority queue and stops weighing on your score.

## Reactivate a silenced finding

Risk acceptance isn't permanent. If the environment changes — or you simply reconsider — **reactivate** a silenced finding to bring it back into your posture, where it counts toward your score again and reappears for remediation.

## Where to go next

<CardGroup cols={2}>
  <Card title="Security Posture" icon="shield-halved" href="/documentation/security-posture/overview">
    Back to the scored dashboard and findings table.
  </Card>

  <Card title="Frameworks & scanning" icon="clipboard-check" href="/documentation/security-posture/frameworks">
    How findings are produced and mapped to frameworks.
  </Card>
</CardGroup>