Skip to main content
ChatGPT connects to NextLevel MCA through developer mode connectors. The sign-in is the same one Claude uses: a code to your work email, pick a workspace, approve what the assistant may do.

​
Add the connector

1

Enable developer mode

In ChatGPT, open Settings → Connectors → Advanced and turn on Developer mode. On Team and Enterprise workspaces an admin may need to allow it.
2

Create the connector

Back in Connectors, choose Create. Name it NextLevel MCA, set the MCP server URL to:
and select OAuth as the authentication method. Leave client id and secret empty; ChatGPT sets itself up automatically.
3

Sign in

Enter your work email on the NextLevel MCA page, type the 6-digit code from your inbox (it works once and expires in 10 minutes), pick a workspace if you have more than one, and approve what ChatGPT may do on the consent screen.
4

Use it in a chat

Start a new chat, open the tools menu, and enable the NextLevel MCA connector. Then ask in plain language; see what you can ask.

​
Doing things needs the matching permission

The consent screen lists every permission with a checkbox, ticked by default. Looking things up (get_deal, get_recommended_lenders, get_statement_analysis and the rest) needs the read permissions; doing things (submit_deal, request_documents, send_merchant_portal, log_offer, move_deal_stage, create_deal, update_deal, set_primary_offer, add_note, create_task) needs the matching write permission on the AI assistants page. A tool whose permission was unticked is still listed but returns a clear message naming it; reconnect and leave it ticked. ChatGPT asks for confirmation before running tools that change things. Keep that on: submit_deal, request_documents and send_merchant_portal send real email or text messages.

​
Revoke access

Revoke from the app under Settings → Developers → Connected apps, or remove the connector in ChatGPT’s settings. Revoking in the app ends the connection immediately; removing it in ChatGPT only deletes ChatGPT’s copy of the access, so do both when offboarding someone. Details are on the Security page.