> For the complete documentation index, see [llms.txt](https://docs.apono.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.apono.io/docs/agent-privilege-guard/apono-agent-privilege-guard-reference/model-context-protocol-mcp-servers.md).

# Model Context Protocol (MCP) servers

Apono supports both **Apono Managed MCPs** and **custom MCPs** for secure agentic access. Apono Managed MCPs are provided and maintained by Apono for supported platforms. Custom MCPs are added and maintained by your organization.

Each MCP exposes a set of tools that an AI agent can call through the Apono Agentic Gateway. Each MCP also has its own intent policy, which governs every tool call made through that MCP. The gateway classifies each action as **Read**, **Create**, **Update**, **Delete**, or **Admin**, then allows it, pauses it for human approval, or blocks it according to the configured guardrail.

***

### Enable and configure an Apono Managed MCP

Enabling an MCP makes its tools available through the Apono Agentic Gateway. It does not grant access to the platform. The agent must also receive access through an applicable agentic access flow created during the [platform-specific setup.](/docs/agent-privilege-guard/set-up-apono-agent-privilege-guard.md)

<figure><img src="/files/zJhbnJpaEVuYia06PUjl" alt="" width="563"><figcaption><p>MCPs tab and panel</p></figcaption></figure>

Follow these steps:

1. On the [**MCPs**](https://app.apono.io/agentic/tools) tab, under **Apono Managed MCPs**, locate the required MCP.
2. On the MCP card, enable it by clicking its toggle.
3. Click the MCP card. The MCP settings panel opens.
4. On the **Intent Policy** tab, select a guardrail for each action category.

{% hint style="info" %}
By default, each MCP allows Read actions, requires human-in-the-loop approval for Create, Update, and Delete actions, and blocks Admin actions.
{% endhint %}

<table><thead><tr><th width="121.0068359375">Guardrail</th><th>Behavior</th></tr></thead><tbody><tr><td><strong>Allow</strong></td><td>The action can proceed without additional approval.</td></tr><tr><td><strong>HIL</strong></td><td><p>The action pauses until the agent’s human operator approves or rejects it.</p><p>Approval can be requested through a supported in-client experience or an external channel such as Slack.</p><p><strong>IMPORTANT</strong>: Human-in-the-loop approval is separate from access-flow approval. Access-flow approval determines whether temporary access is granted. HIL approval determines whether a specific agent action can proceed.</p></td></tr><tr><td><strong>Block</strong></td><td>The action cannot proceed.</td></tr></tbody></table>

5. Click **Save**. The panel closes.

***

### Create a custom MCP server

Follow these steps:

1. On the [**MCPs**](https://app.apono.io/agentic/tools) tab, click **+ Add Custom Tool**. The **Custom MCP** panel opens.
2. On the **Config** tab, enter the **Name**.
3. Select one or more **Resource Types**.
4. Define the **MCP Server Configuration** settings for STDIO or HTTP.

{% tabs %}
{% tab title="STDIO" %}

1. Click **STDIO**.
2. Define the configuration.
   {% endtab %}

{% tab title="HTTP" %}

1. Click **HTTP**.

2. Enter the **Endpoint**.

3. (Optional) Add headers for the MCP:
   1. Under **Headers**, click **+ Add Headers**.
   2. Enter the **Name** and **Value** of any MCP headers.
      {% endtab %}
      {% endtabs %}

4. On the **Intent Policy** tab, select a guardrail for each action category.

{% hint style="info" %}
By default, each MCP allows Read actions, requires human-in-the-loop approval for Create, Update, and Delete actions, and blocks Admin actions.
{% endhint %}

<table><thead><tr><th width="121.0068359375">Guardrail</th><th>Behavior</th></tr></thead><tbody><tr><td><strong>Allow</strong></td><td>The action can proceed without additional approval.</td></tr><tr><td><strong>HIL</strong></td><td><p>The action pauses until the agent’s human operator approves or rejects it.</p><p>Approval can be requested through a supported in-client experience or an external channel such as Slack.</p><p><strong>IMPORTANT</strong>: Human-in-the-loop approval is separate from access-flow approval. Access-flow approval determines whether temporary access is granted. HIL approval determines whether a specific agent action can proceed.</p></td></tr><tr><td><strong>Block</strong></td><td>The action cannot proceed.</td></tr></tbody></table>

6. Click **Save**. The panel closes.


---

# 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, and the optional `goal` query parameter:

```
GET https://docs.apono.io/docs/agent-privilege-guard/apono-agent-privilege-guard-reference/model-context-protocol-mcp-servers.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.