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

# Suggested mitigations

> Actionable remediation steps delivered with every investigation

After every investigation, Wild Moose delivers suggested mitigations - specific, actionable steps your team can take to resolve the incident. These aren't generic suggestions; they're tailored to what the agents actually found during the investigation.

## How mitigations work

When an investigation completes, agents analyze the gathered evidence and recommend remediation steps. Each mitigation includes:

* **What to do** - a clear description of the remediation action
* **Where to do it** - direct links to the relevant tool, dashboard, or pipeline
* **Parameters** - all the context needed to execute the mitigation (service name, environment, threshold values, etc.)

## Types of mitigations

<AccordionGroup>
  <Accordion title="Links to external tools" icon="link" defaultOpen={true}>
    Direct links to runbooks, scaling dashboards, rollback pipelines, or other tools your team uses for remediation. These open in the right context so your team doesn't have to hunt for the right page.
  </Accordion>

  <Accordion title="Executable actions" icon="play">
    Actions that can be triggered directly from the investigation results. These range from simple operations like restarting a service to more complex workflows like rolling back a deployment.
  </Accordion>

  <Accordion title="Always-visible mitigations" icon="eye">
    Some mitigations are relevant regardless of the investigation outcome. You can mark these as `displayAlways` to ensure they're always surfaced - useful for standard operating procedures or escalation paths.
  </Accordion>
</AccordionGroup>

## From suggestion to auto-fix

Wild Moose mitigations exist on a spectrum of automation:

1. **Suggested** - the agent recommends a mitigation and links to the relevant tool
2. **One-click** - the mitigation can be executed directly from the investigation results
3. **Automatic** - for well-understood scenarios, agents can execute mitigations automatically

Your team controls where each mitigation sits on this spectrum. Start with suggestions and gradually increase automation as confidence grows.

<Tip>
  Mitigations are configured as part of each agent's playbook. See [Debugging agents](/features/agents) for details on how to set them up.
</Tip>