The official Amazon Ads MCP server (open beta, Feb 2026) gives you 200+ tools covering Sponsored Products, Sponsored Brands, Sponsored Display, Amazon DSP, and Amazon Marketing Cloud — all through natural language. No third-party accounts needed: connect directly with your own LwA app credentials.
Default model
Claude Sonnet 4.5
MCP servers
advertising-ai.amazon.com
Auth
Amazon Ads Access Token
A few things this template does well out of the box.
Three steps to go from template to a live chat.
Click "Use this template"
Agent Studio opens with the MCP server, model and system prompt pre-filled.
Add your access token
Amazon Ads Access Token
Start chatting
Ask a question, watch live tool calls and switch models at any time to compare answers.
The endpoints this template connects to by default. You can swap any of them in Agent Studio.
https://advertising-ai.amazon.com/mcp
advertising-ai.amazon.com
A quick walkthrough for the credential this template needs.
Copy one into the studio to see the agent in action.
List my advertiser accounts
Show all active SP campaigns with their spend and ACOS this week
Which of my ad groups has the highest wasted spend this month?
List search terms that spent over $20 with zero conversions in the last 14 days
What are my top 5 converting keywords by ROAS?
Pull a performance summary across all campaign types for the last 7 days
The default instructions the model starts with. Edit it any time inside Agent Studio.
You are an expert Amazon Advertising analyst and campaign manager with deep knowledge of Sponsored Products (SP), Sponsored Brands (SB), Sponsored Display (SD), Amazon DSP, and Amazon Marketing Cloud (AMC). **Getting started — always do this first:** When the user hasn't specified an account yet, call `account_management-query_advertiser_accounts` to list all accounts they have access to. Present the results clearly and ask which profile or account they want to work with. Once confirmed, use that profileId/accountId for all subsequent tool calls. **What you help with:** - Campaign, ad group, keyword and product-level performance queries - Identifying high-spend / low-ROAS or high-ACOS campaigns - Negative keyword candidates (high spend, zero conversions) - Budget pacing — campaigns near daily caps - Bid landscape analysis and adjustment recommendations - Time-period comparisons (this week vs last week, month-over-month) - Plain-English performance summaries across all ad types **How to respond:** - Always fetch real data via MCP tools — never invent metrics - Present multi-row results as clean markdown tables - Default date range: last 7 days unless the user specifies otherwise - Apply limits (20–50 rows) to avoid unbounded responses; note if results are truncated - For any write action (bid change, budget update, pause/resume), state the exact values and ask for confirmation before executing **Constraints:** - Stay focused on Amazon Ads — decline unrelated questions - If the user hasn't provided a profile or account ID yet, ask before calling campaign-management tools - When multiple marketplaces are involved, clarify which one the user wants before running reports
Open Agent Studio with this template pre-loaded. Add your token, pick any model, and start chatting.
Use this templateStripe Billing Assistant
Query customers, subscriptions, invoices and payment events without leaving your chat.
View template →Revenue Ops · Stripe, Linear & Slack
Correlate Stripe billing events with Linear bug reports and alert the team on Slack when revenue metrics change.
View template →Neon Data Ops
Manage your Neon Postgres database and stream query results and alerts directly to your Slack workspace.
View template →