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

# Profile in ChatGPT & Claude

> Use your Profile behavioral data inside the AI assistants your team already uses

Bring Profile into the AI assistants your team already uses. Once you connect Profile to ChatGPT or Claude, you can send assessment invites, pull behavioral summaries, and manage tags by just asking, without switching tabs.

<Tip>Profile is designed to work as a ChatGPT app and a Claude connector. Listings in those directories are coming soon. In the meantime you can connect manually using the steps below.</Tip>

<CardGroup cols={2}>
  <Card title="ChatGPT" icon="message-square" href="#chatgpt">
    Connect Profile to ChatGPT's Apps & Connectors.
  </Card>

  <Card title="Claude" icon="bot" href="#claude">
    Add Profile as a connector in Claude.
  </Card>
</CardGroup>

## What you can ask

After connecting, try prompts like these:

```text theme={"system"}
Invite jane@acme.com to take the Profile behavioral assessment. If we already have a "Leadership" tag, attach it.
```

```text theme={"system"}
Pull up Sam Kim's behavioral profile in Profile. Summarize their DISC style and the latest AI debrief in three bullet points I can scan before our 1:1.
```

```text theme={"system"}
Look at everyone tagged "Engineering Leads" in Profile. Based on their behavioral profiles, who's most likely to push back on a new planning process, and how should I frame it?
```

Your assistant handles the call to Profile behind the scenes and answers in plain language, using your actual account data.

## How to connect

<Tabs>
  <Tab title="ChatGPT">
    <Steps>
      <Step title="Open Apps & Connectors">
        In ChatGPT, open the Apps & Connectors screen.
      </Step>

      <Step title="Add Profile">
        Find Profile in the directory once listed, or add a custom app with the URL `https://vanguard.profilebehavior.com/mcp`.
      </Step>

      <Step title="Sign in">
        You'll be sent to Profile to sign in and approve access. Review the permissions shown, then confirm.
      </Step>
    </Steps>
  </Tab>

  <Tab title="Claude">
    <Steps>
      <Step title="Open connector settings">
        In Claude, open **Settings → Connectors**.
      </Step>

      <Step title="Add Profile">
        Find Profile in the directory once listed, or add a custom connector with the URL `https://vanguard.profilebehavior.com/mcp`.
      </Step>

      <Step title="Sign in">
        You'll be sent to Profile to sign in and approve access. Review the permissions shown, then confirm.
      </Step>
    </Steps>
  </Tab>

  <Tab title="Other AI tools">
    Profile speaks the [Model Context Protocol](https://modelcontextprotocol.io), an open standard supported by a growing list of AI tools. If your assistant can add a custom MCP server, point it at `https://vanguard.profilebehavior.com/mcp` and sign in when prompted.
  </Tab>
</Tabs>

## What your AI assistant can access

When you connect Profile, you're signing in with your Profile account and approving a specific set of permissions. Your AI assistant only gets what you explicitly allow, and only for the account you signed in with.

<Check>You'll see a consent screen listing exactly what the assistant is asking for before anything is connected.</Check>

Common permissions include:

* **View profiles and behavioral data.** Read names, scores, and debriefs for people on your account.
* **Send invites and manage profiles.** Invite new people to take the assessment, update profile details.
* **Manage tags.** Create and organize groups of people on your account.

You can deny individual permissions at consent time, and you can disconnect the assistant entirely at any time.

## Managing your connections

To see every AI assistant or third-party app you've connected to Profile:

<Steps>
  <Step title="Sign in to Profile">
    Open Profile in your browser and sign in.
  </Step>

  <Step title="Open your user profile">
    Go to the **User Profile** page.
  </Step>

  <Step title="Find Connected apps">
    Scroll to the **Connected apps** section.
  </Step>
</Steps>

From there you can review what's connected and revoke any connection instantly. Once revoked, that assistant loses access immediately, and you can always reconnect it later if you change your mind.

## Privacy and data handling

Connecting an AI assistant doesn't give it unrestricted access to your Profile account. The assistant can only perform the actions you approved, scoped to the account you signed in with. If you belong to multiple Profile accounts, you'll need to connect each one separately.

For more on how Profile handles behavioral data overall, see our [Data Privacy](/data-privacy) guide.

## Frequently asked questions

<AccordionGroup>
  <Accordion title="Do I need a special plan?" icon="credit-card">
    No. If your account has API access, your team can connect AI assistants. Contact your account owner if you're not sure.
  </Accordion>

  <Accordion title="Can my assistant see other Profile accounts I belong to?" icon="shield">
    No. A connected assistant is tied to the specific account you signed into during the connection flow. To give an assistant access to a different account, sign into that account in Profile first, then connect from there.
  </Accordion>

  <Accordion title="What does the assistant do with my data?" icon="database">
    That depends on the assistant. ChatGPT, Claude, and other AI tools each have their own data policies covering what they do with information during a conversation. Consult the policy of the assistant you're connecting.
  </Accordion>

  <Accordion title="Can I undo an action the assistant took?" icon="rotate-ccw">
    Actions like sending an invite or deleting a tag are real. Review the assistant's plan before confirming anything destructive. Profile logs actions made by connected assistants so your team has a record of what happened.
  </Accordion>

  <Accordion title="What happens when I revoke access?" icon="ban">
    The assistant loses access immediately. Any open chat threads referencing Profile data will still show that data (it's already in the conversation), but the assistant can't make new requests to Profile until you reconnect.
  </Accordion>
</AccordionGroup>