Add one MCP server to Gemini and it answers from your Seller Central, Vendor Central and Amazon Ads data. Switch on write access and it can update listings, prices and campaigns too, after you confirm each change.
Three listings dropped the most. Each can win back the offer above your margin floor:
Bamboo cutting board XL · US · $36.49 → $34.99
Previewed as a dry run · margin after fees 21.4%Ask Gemini
Every Google product that speaks the Model Context Protocol reaches DataDoe at the same URL. Only the label in the menu and the way you sign in change.
Custom app
Custom MCP server data store
MCP server
gemini mcp addMCP server
Behind it: Seller Central, Vendor Central and Amazon Ads, already joined.
Each Google surface adds MCP servers its own way. Choose yours and follow four steps. Connecting Amazon to DataDoe takes about five minutes and only happens once.
For owners and operators who want answers on the web and on their phone.
You need
Google adds custom apps on the web only, and not yet for work or school accounts. Teams on Google Workspace use Gemini Enterprise.
Sign in to Seller Central, Vendor Central and Amazon Ads through Amazon’s consent screens. DataDoe never sees a password.
On gemini.google.com, open Settings → Connected Apps and, under Custom apps, choose Add a custom app.
Enter https://mcp.datadoe.com/mcp/v1, click Next and sign in to DataDoe when the browser asks.
In any chat, type @ and pick DataDoe. Once added on the web, it shows up in the Gemini mobile app too.
For teams on Google Workspace who want one shared connection that an admin controls.
You need
Custom MCP servers in Gemini Enterprise are a Google Preview feature, so menu names can still move.
In the Gemini Enterprise console, an admin creates a data store of type Custom MCP Server with the URL https://mcp.datadoe.com/mcp/v1.
Fill in the OAuth fields Google asks for and sign in to DataDoe once. If you need client details from us, write to support.
Google switches no tools on by default. Enable the DataDoe read tools first, and the write tools only for teams that need them.
Connect the data store to your Gemini Enterprise app, and people can use DataDoe in chat and in agents.
For developers who script reports, checks and small tools from the terminal.
You need
Since 18 June 2026, Google AI Pro and Ultra subscribers use Antigravity CLI in place of Gemini CLI. Paid API keys and Code Assist Standard or Enterprise keep Gemini CLI.
In DataDoe, open Integrations → MCP and create a key for this project.
From your project root, run:
gemini mcp add --transport http --scope project -H "datadoe-mcp-key: YOUR_DATADOE_MCP_KEY" datadoe "https://mcp.datadoe.com/mcp/v1"
Open a new Gemini CLI session, then run gemini mcp list. A disabled DataDoe entry usually means the folder is not trusted yet.
Scheduled reports, scripts and the fallback config are in the Gemini CLI guide for Amazon.
For builders who want an agent IDE to write internal Amazon tools against real data.
You need
Antigravity uses the key serverUrl for remote servers, not url.
In the agent side panel, open … → MCP Servers → Manage MCP Servers → View raw config.
Add this entry and keep any servers already there:
{
"mcpServers": {
"datadoe": {
"serverUrl": "https://mcp.datadoe.com/mcp/v1"
}
}
}In Agent settings, open Customizations and click Authenticate next to DataDoe. Sign in, then paste the code you get back.
Ask: “Use DataDoe to list my connected Amazon accounts.”
Seller Central, Vendor Central and Amazon Ads arrive in one schema of 130+ tables, keyed by SKU and ASIN. Reading works from the first day. Writing stays off until you switch on each Action type.
Gemini can read
Gemini can change, once you allow it
Gemini can read
Gemini can change, once you allow it
Invoices stay read-only.
Gemini can read
Gemini can change, once you allow it
The Gemini app asks before every write. DataDoe adds its own switches long before that question comes up, so a stray prompt can’t touch a bid or a price.
Every Action type in DataDoe starts off. You turn on prices, listings or ads one at a time in Settings → Actions.
Each Amazon account stays read-only until you change it on the Accounts page.
Ask Gemini to dry-run first, and DataDoe validates the payload while keeping it off Amazon.
Google’s Gemini app asks for manual confirmation before any write. Nothing goes out until you press Confirm.
Every Action lands in DataDoe’s Action history: what changed, when, and on which account.
The DataDoe connection is the same on every surface. What changes is how deep the question goes, and whether you need the answer once or every week.
One table, one answer. A fast model on your phone is enough.
Several sources joined and weighed. Pick Pro, and Deep Think for the hardest questions if your plan has it.
Write it once as a script or a small tool, then let it run on fresh data every time.
Gemini Enterprise gives everyone the same DataDoe connection, with an admin deciding which tools are on. Agents built there can use it too.
Which models you see depends on your Google plan. Switching models never means reconnecting DataDoe.
In the Gemini app, the @ mention tells Gemini where to look before it starts guessing. Copy a prompt, swap in your numbers and send it.
In Gemini CLI and Antigravity, skip the @ and say “use DataDoe” instead.
This page covers every Google surface. The terminal has a guide of its own, and every other AI client has its own hub.
Use another AI? Each one has a hub.
MCP, the Model Context Protocol, is the open standard Gemini uses to call outside tools. Google supports it as custom apps in the Gemini app, as custom MCP server data stores in Gemini Enterprise, and as MCP servers in Gemini CLI and Antigravity. DataDoe is an MCP server for Amazon at https://mcp.datadoe.com/mcp/v1.
On gemini.google.com, open Settings, then Connected Apps, and under Custom apps choose Add a custom app. Paste https://mcp.datadoe.com/mcp/v1, click Next and sign in to DataDoe. Google currently offers this to personal accounts in the US, aged 18 or over, with Keep Activity on. Then type @ in a chat and pick DataDoe.
Not until you say so. Writes go through DataDoe Actions, which ship switched off; you enable them type by type, from listings and A+ Content to FBM shipments and ad campaigns, and prices have a type of their own. In the Gemini app, when Gemini wants to write, it shows a confirmation and waits. A dry run can come first, and DataDoe records each change in its Action history.
Not in the consumer Gemini app yet, because Google limits custom apps there to personal accounts. Workspace teams use Gemini Enterprise, where an admin adds DataDoe as a custom MCP server data store, a feature Google still lists as Preview.
Yes, with a paid Gemini API key or a Gemini Code Assist Standard or Enterprise licence. On 18 June 2026 Google moved Google AI Pro and Ultra subscribers to Antigravity CLI. Both setups are in the Gemini CLI guide for Amazon.
Use Flash for quick lookups that sit in one report. Use Pro when the answer joins several sources, such as fees, ads and costs, and Deep Think for the hardest questions if your plan includes it. The DataDoe connection stays the same whichever model you pick.
No, the two are billed separately. DataDoe costs $97 per month once the 14-day free trial ends, and your Seller Central, Vendor Central and Amazon Ads accounts all sit on that one plan. MCP use is part of the plan and counts against its AI Tokens. Your Gemini or Google plan is billed by Google.
DataDoe sends Gemini only the rows a question needs, read through Amazon's official APIs. What Google keeps from the chat follows your Gemini activity settings or your Workspace terms. Google also notes that it does not run third-party MCP servers, so DataDoe secures its own side.
We'll connect a demo account to Gemini, ask it real questions about Amazon and show you how a confirmed change goes through.
Book a demoConnect Seller Central, Vendor Central and Amazon Ads once. Every Google surface that speaks MCP can then read them — and change them when you confirm.
We use cookies to improve your experience and analyze traffic. By clicking "Accept", you agree to our use of cookies. Read the Cookie Policy.