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

# AI coding agents

> Connect your own AI coding agents to your Amnify security data.

The agent connector lets your own AI coding agents read your Amnify security data and act on it, inside the same permission boundary as the product. It speaks the Model Context Protocol, so any client that supports MCP can connect.

## Connecting

<Steps>
  <Step title="Point your agent at Amnify">
    Add Amnify as a connector in your AI coding agent, using the endpoint below.

    ```
    https://mcp.amnify.ai/mcp
    ```
  </Step>

  <Step title="Sign in and choose the organization">
    You authenticate as yourself and pick the organization the connection is for.
  </Step>

  <Step title="Approve the access">
    The consent screen names the client, the organization, and what is being requested, in plain language.
  </Step>
</Steps>

## What a token can reach

A token is bound to exactly one organization at the moment it is issued. It cannot be switched to another one later, and connecting a second organization requires a separate authorization.

Everything a client can reach is filtered by that organization, using the same isolation rules as the web application. See [Data security](/platform/data-security).

## What it is useful for

<CardGroup cols={2}>
  <Card title="Ask in context" icon="comment">
    Query findings, scans and context from wherever you already work, without switching tools.
  </Card>

  <Card title="Bring security into your own agents" icon="sitemap">
    Let a coding agent check the security state of a service before it changes it.
  </Card>

  <Card title="Trigger work" icon="play">
    Start a scan or pull the current state of a scope as part of your own automation.
  </Card>

  <Card title="Report" icon="chart-column">
    Pull the current picture into your own reporting rather than exporting by hand.
  </Card>
</CardGroup>

## Revoking

Revoke an agent's access at any time from your organization settings. Access ends immediately.