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

# Connect your AI tool to the Operata Trust Center

> Connect Claude, ChatGPT, Cursor, or another AI tool to the Operata Trust Center over the Model Context Protocol, so you can ask about Operata's security posture and pull documents during a security review.

Connect your AI tool to the [Operata Trust Center](https://trust.operata.com) so it can answer security-review questions and fetch documents for you. Ask "What certifications does Operata have?" and your AI tool searches the Trust Center and answers.

The Trust Center runs on SafeBase. SafeBase's Trust Center connection uses the Model Context Protocol (MCP). It's read-only, and it shows only what your Trust Center access already lets you see. It's separate from the [Operata MCP Server](https://docs.operata.com/docs/guides-mcp-intro.md), which queries your contact center data.

## Before you start

* **Approved Trust Center access.** Go to [trust.operata.com](https://trust.operata.com), select **Get access**, and submit the request. When it's approved, click the confirmation link in the email you receive. Connecting your AI tool doesn't grant access on its own.
* **A supported AI tool:** Claude (claude.ai, Claude Desktop, or Claude Code), ChatGPT (Plus, Pro, Business, Enterprise, or Edu), Cursor, Codex CLI, Windsurf, or VS Code.
* **Admin enablement, on some plans.** On Claude Team and Enterprise, a workspace admin may need to enable custom connectors (**Organization settings** > **Connectors**). On ChatGPT Team, Enterprise, and Edu, a workspace admin may need to enable Developer mode.

You don't need a token, API key, or password. You sign in with the same email-based access you use to view the Trust Center.

## Steps

Every AI tool uses the same server URL:

```text theme={null}
https://app.safebase.io/api/trust-center/mcp
```

The URL is the same for every vendor on SafeBase. The Trust Center you're signed in to when you approve the connection decides which vendor you reach, so sign in to the Operata Trust Center first.

1. Sign in to [trust.operata.com](https://trust.operata.com) in your browser.
2. Add the server to your AI tool, using the tab for your tool below.
3. When your browser opens the SafeBase authorization screen, check that you started the request and that the destination is the Operata Trust Center, then approve. The approval lasts until your access is revoked or the session expires.

> [tab] Claude
>
> In claude.ai or Claude Desktop, go to **Customize** > **Connectors** and add a custom connector. Enter the server URL and sign in. To use it in a chat, turn it on from the **+** button.

> [tab] ChatGPT
>
> Go to **Settings** > **Connectors**. Under **Advanced**, turn on **Developer mode**. Select **Create**, enter the server URL, and sign in.

> [tab] Claude Code
>
> Add the server:
>
> ```bash theme={null}
> claude mcp add --transport http trust-center https://app.safebase.io/api/trust-center/mcp
> ```
>
> Then run `/mcp` inside Claude Code and sign in through your browser.

> [tab] Cursor
>
> Add a `trust-center` entry under `mcpServers` in `~/.cursor/mcp.json`, or in `.cursor/mcp.json` for a single project:
>
> ```json ~/.cursor/mcp.json theme={null}
> {
>   "mcpServers": {
>     "trust-center": {
>       "url": "https://app.safebase.io/api/trust-center/mcp"
>     }
>   }
> }
> ```
>
> Restart Cursor and sign in when prompted.

> [tab] Codex CLI
>
> Add the server, then sign in on first use:
>
> ```bash theme={null}
> codex mcp add trust-center --url https://app.safebase.io/api/trust-center/mcp
> ```

> [tab] Windsurf
>
> Add a `trust-center` entry under `mcpServers` in `~/.codeium/windsurf/mcp_config.json`:
>
> ```json ~/.codeium/windsurf/mcp_config.json theme={null}
> {
>   "mcpServers": {
>     "trust-center": {
>       "serverUrl": "https://app.safebase.io/api/trust-center/mcp"
>     }
>   }
> }
> ```
>
> Sign in when prompted.

> [tab] VS Code
>
> Run **MCP: Add Server** from the Command Palette, or add an entry under `servers` in `.vscode/mcp.json`:
>
> ```json .vscode/mcp.json theme={null}
> {
>   "servers": {
>     "trust-center": {
>       "type": "http",
>       "url": "https://app.safebase.io/api/trust-center/mcp"
>     }
>   }
> }
> ```
>
> Sign in when prompted.

## Verify

Ask your AI tool:

> What Trust Centers do I currently have access to?

The answer lists Operata and flags it as the Trust Center your session belongs to. Then try a security-review question:

* "What certifications does Operata have?"
* "Show me Operata's subprocessors."
* "Does Operata have a recent penetration test report I can download?"
* "Has anything changed with Operata's compliance status recently?"

You don't need to name a tool — your AI tool picks one. The connection provides four read-only tools:

| Tool | What it does |
| - | - |
| `tc_access` | Lists every Trust Center you can access, with organization names and public URLs, and flags the one your session belongs to. |
| `tc_explore` | Browses a Trust Center's products, pages, and content groups — certifications, reports, policies, and questionnaires — from overview down to item detail. |
| `tc_search` | Searches documents, knowledge base entries, content items, and published updates, or lists every item of one content type. |
| `tc_document_download` | Returns a time-limited, signed download URL for a document, using an ID from `tc_search` or `tc_explore`. |

If a search comes back empty or unclear, ask your AI tool to browse the Trust Center's structure instead.

## Troubleshooting

| Symptom | Fix |
| - | - |
| The sign-in screen asks you to request access. | You aren't signed in to the Operata Trust Center. Request access at [trust.operata.com](https://trust.operata.com), click the confirmation link in your email, then reconnect. |
| Your AI tool answers about a different vendor, or Operata is missing. | Ask your AI tool to list the Trust Centers you can access. If Operata isn't the current one, sign in to [trust.operata.com](https://trust.operata.com) and approve the connection again. Each vendor needs its own connection. |
| A download link no longer works. | Download links expire. Ask your AI tool for a new link. |
| Claude or ChatGPT has no option to add a custom connector. | Your workspace admin needs to enable custom connectors (Claude) or Developer mode (ChatGPT). |
| You want to disconnect. | Remove the connector or server entry from your AI tool's settings or config file. Your Trust Center access stays as it is. |

For anything the Trust Center doesn't answer, contact your Operata account team.

## Related

* [Operata Trust Center](https://trust.operata.com) — request access and browse documents in the portal.
* [SafeBase Trust Center MCP user guide](https://help.safebase.io/en/articles/16203034-safebase-vendor-trust-center-mcp-user-guide-early-preview) — SafeBase's own setup guide.
* [MCP overview](https://docs.operata.com/docs/guides-mcp-intro.md) — the Operata MCP Server, for querying your Operata data.


## Related topics

- [Connect your AI tool to the Operata docs](https://docs.operata.com/docs/guides-docs-mcp.md)
- [Diagnose voice incidents with Amazon Bedrock](https://docs.operata.com/docs/guides-mcp-bedrock-agentcore.md)
- [Connect ChatGPT to the Operata MCP server](https://docs.operata.com/docs/guides-mcp-chatgpt.md)