Connecting GBPly to Your AI Assistant
Go to AI Connector (Claude, ChatGPT) →If you use an AI assistant such as Claude, ChatGPT or Cursor, you can connect it to GBPly so it can read your profiles, reviews, posts and reports, draft replies and posts, and ask you to approve changes. The assistant works as you: it sees only the workspaces and profiles you can see, and it can never publish, change Google or spend credits without your approval in GBPly.
Before You Start
You need a GBPly account with access to at least one workspace, and an AI assistant that supports custom connectors (also called MCP servers or remote MCP). Claude (web and desktop), ChatGPT (with connectors or developer mode enabled on your plan), Cursor, VS Code and the MCP Inspector all support this.
The connector address is the same for everyone:
https://mcp.gbply.net/v1You don't need an API key. When you add the connector, your assistant opens a GBPly sign-in and approval screen where you choose Approve or Deny.
Connecting Claude
In Claude, open Settings → Connectors and choose Add custom connector.
Name it GBPly and paste the connector address above. Leave any client ID and secret fields empty.
Click Connect. A GBPly window opens; sign in if asked, check the screen says it's your account, then click Approve.
Back in a Claude chat, enable GBPly in the tools menu and try: "List my GBPly business profiles."
Connecting ChatGPT
In ChatGPT, open Settings → Apps & Connectors (on some plans you first turn on Developer mode under Advanced settings), then choose Create or Add connector.
Enter GBPly as the name, paste the connector address, and choose OAuth as the authentication method.
Approve the GBPly sign-in window when it appears, then select GBPly in a new chat and ask it about your profiles.
Connecting Cursor, VS Code and Other Assistants
Any assistant that supports remote MCP servers over HTTP with OAuth sign-in can connect. Add a server with the connector address; for example, in Cursor's mcp.json:
{
"mcpServers": {
"gbply": { "url": "https://mcp.gbply.net/v1" }
}
}The assistant registers itself automatically and opens the GBPly approval screen the first time it connects.
What the Assistant Can Do
Read: your workspaces and profiles, profile details and completeness, performance insights (Google's data runs 2–3 days behind), monthly search terms, reviews and rating stats, posts, photos, Q&A, Profile Guard changes, Rank Tracker scans and results, citation scans, analysis reports, and credit balances.
Draft (nothing is saved or posted): AI review replies and AI Q&A suggestions; cost previews for Rank Tracker scans.
Ask you to approve: posting, editing or deleting a review reply; publishing, scheduling, editing or deleting a post; publishing a saved FAQ; editing your description, phone, website, hours, special hours, services or categories; accepting, dismissing or reverting a Profile Guard change; running a Rank Tracker scan, a citation scan or an industry analysis.
Directly: sync a profile from Google (a few times an hour at most).
The connector doesn't handle billing, plan changes, workspace members, connecting Google accounts, review-request contact lists, account deletion or data export. Use GBPly for those.
Approving Requests
When the assistant wants to change something, it creates a request and gives you a link. Open it (or go to Assistant requests in the menu) while signed in to GBPly.
Each request shows exactly what will happen: the reply or post text, the image, the current and new value of a profile field, or the credit cost. Choose Approve or Reject, or Approve all when several are waiting (for example, a week of posts).
Requests expire after 15 minutes and can only run once. When you approve, GBPly checks again that you still have permission, have enough credits and are within your plan limits and daily post cap. The usual safeguards still apply: posts that fail the post policy check are kept as flagged, not published.
Your assistant can check the result afterwards. It should never tell you a change was made until GBPly reports it as done.
If you didn't ask your assistant for a change, just reject the request. Nothing happens to rejected or expired requests.
Permissions, Limits and Safety
The assistant gets exactly your permissions in each workspace. Viewers can read and draft but can't create requests; owners, admins and editors can. Staff and admin features are never available through the connector, and demo profiles are read-only.
If a workspace requires two-factor authentication, you need to sign in with 2FA to approve requests there.
Usage is rate-limited per person (reads, AI drafts, syncs and requests each have their own limit). If you hit one, the assistant is told how many seconds to wait.
Contact lists, sign-in tokens, billing details and other people's email addresses are never sent to the assistant.
Switching Off or Disconnecting
Workspace owners can switch AI connectors off for a workspace in Workspace Settings → AI connectors. This takes effect immediately for every connected assistant.
The same card lists the assistants you've connected. Click Disconnect to remove one; it will need your approval again to reconnect. You can also see recent assistant activity in that workspace.
Example Prompts
Small business: "Which of my reviews haven't been answered? Draft friendly replies and send them for my approval." · "Write three posts for next week about our spring offer and schedule them for 9am." · "How did my calls and website clicks change last month?"
Agency: "Across all profiles in my agency workspace, list any with unanswered 1–2 star reviews." · "Preview the cost of a 7×7 rank scan for 'emergency plumber' for each client, then request scans for the three with the lowest scores." · "Summarise Profile Guard changes Google made this week and suggest which to revert."
