Trust Center Visitor AI Connector

Time to complete: 5 minutes
Prerequisites: A work email address.

Overview

The Trust Center Visitor AI Connector lets you plug your AI assistant into Conveyor-hosted Trust Centers you review. Connect Claude, ChatGPT, or any other MCP-compatible assistant and you can search for vendors on Conveyor, browse their security documents, download SOC 2 reports, ask compliance questions, and request access to gated content — all through natural conversation.

Under the hood, the Visitor AI Connector is a hosted MCP server purpose-built for buyer workflows. MCP is an open standard that gives AI models a structured way to call external tools — similar to an API, but for AI assistants. Connect once and it works with any assistant that supports MCP: no custom integration, no API scripting, and setup takes seconds.

Note: The vendor must opt in to making their Trust Center available via MCP.

📘
What can I do with it?
Here are a few examples of things you can ask your AI assistant once connected:

"Find Acme Corp's trust center and show me their available documents."

"Download Acme's SOC 2 Type II report."

"Does Acme encrypt data at rest?"

"Request access to Acme's gated security documents."

See Using the Visitor AI Connector for full workflows.

Available Tools

Once connected, your AI assistant automatically discovers all available tools. These tools let you interact with any Conveyor-powered Trust Center — searching for vendors, browsing documents, asking security questions, and more — all through natural language.

Because the Visitor AI Connector speaks standard MCP, it isn't limited to chat assistants — any system that can run an MCP client can connect, including AI platforms like Notion AI and Atlassian Rovo.

What you can see and do depends on your access level at each trust center. Public content is always available, but gated documents and detailed responses may require vendor approval. When you request access, the vendor is notified and can approve, deny, or require an NDA — the same workflow as visiting the trust center directly in a browser.

Authentication

Mode	How it works	Best for
OAuth 2.1	Click Connect in your AI client, sign in with your work email, and you're authenticated. Your identity is verified and carried through to every trust center you visit.	Claude, ChatGPT, any modern MCP client

📘
How does authentication work?
The Visitor AI Connector uses OAuth to verify your identity. When you connect, you sign in once and your email is used to determine your access level at each Trust Center — the same way it works when you visit a Trust Center in your browser. If a vendor has already approved your email, your assistant gets the same access automatically.

Ready to connect? Follow Connecting Your AI Client for step-by-step setup in Claude, ChatGPT, and other clients.

Common Questions

Q: Do I need to self-host anything?
A: No. Conveyor hosts the Visitor AI Connector. Just point your client to the MCP server URL and authenticate via OAuth.

Q: Can the AI assistant access Trust Centers I haven't been approved for?
A: No. It can find and view public content on any Conveyor-powered Trust Center, just as you could in a browser. But gated content requires vendor approval. Your AI assistant can request access on your behalf, but can't bypass access controls.

Q: What are the different access levels?
A: When you visit a Trust Center, your status will be one of: authorized (full access, approved by the vendor), requires NDA (NDA signing needed to unlock gated content), pending approval (your access request is awaiting vendor review), or no access (default — you can see public content only).

Q: How do I request access to gated documents?
A: Just ask your AI assistant to request access. It will submit an access request to the vendor on your behalf. You'll need to wait for vendor approval before gated content becomes available. You can ask your assistant to check your status at any time.

Q: Can I ask security questions directly?
A: Yes. The Visitor AI Connector includes a tool that lets your assistant ask security questions to a vendor's Trust Center — the same Q&A experience available on the Trust Center website. Responses are generated from the vendor's approved knowledge base.

Q: Does the vendor see my questions?
A: Your AI conversation stays between you and your AI provider. However, questions submitted through the Q&A tool are sent to the vendor's Trust Center and may be visible to the vendor, just as they would be if you asked the same question on their Trust Center website.

Q: Can I download documents?
A: Yes. Your assistant can download any document you have access to, including PDFs. PDF text is automatically extracted so your assistant can read and summarize the content inline — no need to open a separate file.

Q: How do I disconnect?
A: Remove the Visitor AI Connector from your AI client's MCP settings. Your access to individual Trust Centers is managed by each vendor independently.

What's Next?

Connecting Your AI Client — step-by-step setup for Claude, ChatGPT, and other clients.
Using the Visitor AI Connector — example workflows for vendor security reviews, document collection, and compliance Q&A.

Need help? Visit the Integration Issues guide or contact [email protected].