BrandOps9

Help · Guide

Connect an AI agent (Claude, ChatGPT, MCP)

Connect Claude, ChatGPT or any MCP-capable agent to BrandOps9 so it can read your workspaces and draft, approve and schedule content for you.

What this is

Agent access exposes BrandOps9 as an MCP (Model Context Protocol) server. Once connected, an AI agent can work inside your account: read your campaigns, content, credits, schedule and learning results and, if you allow it, create a brand workspace from a website URL, generate and revise posts and collateral, approve them and place them on the calendar.

Everything the agent does is scoped to your own workspaces and written to the audit log. It can never do more than you could do in the app. Agent access is available on the Growth, Scale and Agency plans.

Open the connect screen

Open the Setup menu in the top navigation and choose Agent access. That screen explains when to use one-click connect versus a token, shows your MCP endpoint URL and step-by-step instructions per client, and is where you create and revoke access tokens.

Connect Claude or ChatGPT (one click)

In Claude (claude.ai) or ChatGPT, open Settings → Connectors and add a custom connector using your MCP endpoint URL. Leave any OAuth Client ID / Secret fields empty. When you connect, BrandOps9 opens a sign-in and approval page where you choose exactly what the agent may do (read, create, or approve & schedule). Approve, and you're connected. There is no token to copy.

  • Read: see workspaces, content, credits, schedule and learning.
  • Read + Create: also create workspaces and generate campaigns, posts and collateral as drafts (uses credits).
  • Read + Create + Approve: also approve content and schedule it to publish.

Connect Claude Code or a custom client (token)

For Claude Code or any client that authenticates with a header, create a Personal Access Token on the Agent access screen (pick its scope). The screen gives you a ready-to-run command with your new token already filled in. Copy, paste in your terminal, done.

Copy the token when it's shown; it isn't displayed again. You can revoke any token or connection anytime from the same screen.

Staying in control

You choose the scope when you connect, so a read-only connection can never change anything. Every agent action runs through the same role, credit and content-safety checks as the app, and publishing still follows your normal rules: scheduled posts go out via your connected accounts at their planned time. Revoke a connection at any time from Setup → Agent access; it stops working immediately.