Read example
Compare channel performance
“Compare Amazon and Noon sales over a selected date range. Which channel had the best ROAS?”
Amazon Seller CentralNoon SellerAds source
MCPOne connection layer for selected business sources
Ask your AI about sales, ads, and inventory — across your connected sources.
SellerPort MCP connects documented commerce, advertising and analytics sources to Claude and supported MCP clients. Start with source data — not another export.
Get started (opens in a new tab)Visit mcp.sellerport.io (opens in a new tab)Explore supported sources
- 8 published active sources
- 52 documented read tools
- Source and token guidance
Illustrative data flowsample question, not a product screen
READ EXAMPLECONNECTED SOURCE TYPES
CommerceAdvertisingAnalytics
SellerPort MCPSelected business tools
“Compare selected source performance for the last 30 days. Where should I investigate first?”
Illustrative read querySource fit
Bring the sources you use into one MCP workspace.
Check source status and documentation before you connect. Every source remains clearly named — no generic “all your data” promise.
Commerce
Advertising
How it works
Your AI client asks. SellerPort connects the right business tools.
MCP is a standard that lets an AI client call external tools. SellerPort MCP sits between documented sources and your configured client, so a question can start with selected business data instead of pasted reports.
Conceptual connection diagram — not a product screen
1 · CONNECTSelected source
Authorize the documented source path.
2 · CONTROLSellerPort MCP
Published tools and access model.
3 · ASKAI client
Use the setup documented for your client.
First workflow
Start with one source, one client, one read question.
Step 01: Connect a source
SellerPort documents approval-based paths for Amazon, Google and Meta. Noon is different: its guide describes a one-time service-account JSON or private-key step.
Read source setup (opens in a new tab)Step 02: Configure your client
Choose the client setup that matches your workflow. Follow the documented path rather than assuming every client connects in the same way.
View client guides (opens in a new tab)Step 03: Ask a read question
Begin with a bounded question that names the sources and timeframe you need. Review current tool detail before allowing account changes.
Read tool reference (opens in a new tab)
Practical prompts
Questions that start with the source data.
Examples help the right team recognize their workflow. They do not show a customer result, an automated recommendation or a promise of a particular answer.
Read example
Review inventory context
“Show documented Amazon inventory context alongside recent sales performance for the selected period.”
Amazon Seller Central
Read example
Investigate campaign context
“Review selected campaign performance alongside connected business data before deciding what to investigate.”
Amazon AdsGoogle AdsMeta Ads
Clear action boundary
61 documented business tools. Start with the 52 that read.
SellerPort publicly lists 52 read tools and 9 write tools across active sources. Those are different responsibilities — the page treats them differently too.
PUBLISHED TOOL SURFACE
52 read tools
Use source data to investigate commerce, advertising and analytics questions before deciding what to do next.
- Commerce source data
- Advertising performance context
- Analytics and search signals
DOCUMENTED CHANGE SURFACE
9 write tools
A limited published set can affect selected prices, listing details, inventory, shipments, campaign settings, bids and Vendor PO responses.
Review before use. Understand the tool and workspace controls before enabling account changes.
Read security and write detail (opens in a new tab)Client fit
Choose the setup documented for your workflow.
Client support is not one-size-fits-all. Start with the public guide that matches your environment and check current compatibility before connecting.
| AI client | Public status | Published setup direction | Guide |
|---|---|---|---|
| Claude Desktop | Documented setup | Local client guide | Open guide (opens in a new tab) |
| Claude Code | Documented setup | Public client documentation | Open guide (opens in a new tab) |
| Cursor | Documented setup | Public client documentation | Open guide (opens in a new tab) |
| Windsurf | Documented setup | Public client documentation | Open guide (opens in a new tab) |
| ChatGPT | Beta / current check | SellerPort describes custom actions support | Check status (opens in a new tab) |
Connection model
A connection model you can inspect.
These are SellerPort’s published security and privacy statements. Read the current documentation for the details that matter to your team.
Server-side credentials
SellerPort states that provider credentials are held server-side while the AI client receives an MCP token.
Security detail (opens in a new tab)Tokens and revocation
Public documentation describes MCP tokens and guidance for disconnecting an account or revoking a token.
Token guidance (opens in a new tab)Published encryption model
SellerPort’s security page says credentials use AES-256-GCM encryption and MCP tokens are hashed.
Read the statement (opens in a new tab)Data behavior
Its privacy page says data is requested on demand with short-lived report caching.
Privacy detail (opens in a new tab)Noon connection note
SellerPort documents a one-time Noon service-account JSON or private-key step. Review this before you connect.
Read source setup (opens in a new tab)Product access
Start with the sources you need.
Review source and client setup, then create access when you are ready. Current commercial terms should be confirmed in the product before you connect an account.
Get started (opens in a new tab)Questions, clearly answered
Evaluate the connection before you make it.
Source state, credential handling and write boundaries are more useful than generic AI promises.
What is SellerPort MCP?
SellerPort MCP is presented as a layer that lets an AI client call tools for connected commerce, advertising and analytics sources instead of relying on exported spreadsheets.
Read product overview (opens in a new tab)Which sources are active today?
SellerPort publicly lists Amazon Seller Central, Amazon Ads, Amazon Vendor Central, Noon Seller, Google Ads, Meta Ads, Google Analytics 4 and Search Console as active. Shopify order and product MCP tools are marked coming soon.
Review integrations (opens in a new tab)Do I paste provider API keys?
SellerPort documents approval-based source connections for Amazon, Google and Meta. Its Noon guide describes a one-time service-account JSON or private-key step.
Read source setup (opens in a new tab)Can an AI client change connected accounts?
SellerPort publishes 61 business tools: 52 read tools and 9 write tools. Read the current tool reference and security detail before enabling change access.
View tool reference (opens in a new tab)Which AI clients can I use?
SellerPort publicly documents setup for Claude Desktop, Claude Code, Cursor and Windsurf. It presents ChatGPT as beta/custom-actions support; check current compatibility before connecting.
See client guidance (opens in a new tab)How are source connections and tokens revoked?
SellerPort’s public documentation describes disconnecting an account or revoking an MCP token. Verify the current workflow in your workspace before relying on a particular process.
Read token guidance (opens in a new tab)Your next source question
Start with the source question you need answered.
Review supported sources and client setup, then create access when you are ready.