Chat assistants
Connect Kuudo as an MCP server inside the chat your team already uses. Ask for campaigns, listings, or reports in plain language. Answers come back grounded in live Amazon data.
MCP server for the Amazon Advertising API
An MCP (Model Context Protocol) server for the Amazon Advertising API: adjust bids, cut wasted spend, shift budget, pull the report — from ChatGPT, Claude, Cursor, or whatever AI tool your team already uses. The whole Amazon Ads stack, run by your agents instead of by hand.
What your agents can do
Complete coverage: 711 Amazon Ads API operations across 53 services, every one a job your agents run directly. From a single bid change to full-funnel measurement — sponsored campaigns, DSP, AMC, the new unified Ads API v1 with its beta operations, attribution, and reporting included — no console required.
Campaigns, ad groups, keywords, product targets, and bids — Sponsored Products run end to end by your AI.
Sponsored Brands v3 and v4: campaigns, creatives, and headline placements, managed by your AI.
Audience and contextual Sponsored Display campaigns, targeting, and creatives, run by your AI.
Streaming TV campaigns and creatives, launched and optimized by your AI.
Programmatic line items, advertisers, audiences, conversions, and measurement across the demand-side platform.
The unified campaign-management API: create, update, and pace campaigns across every ad program at once.
Amazon's new unified Ads API: cross-program operations plus per-program surfaces for Sponsored Products, Brands, Display, TV, and DSP — v1 beta operations included.
Keywords, product targets, negatives, and bid recommendations created and tuned by your AI.
Creative assets, moderation, and pre-moderation handled by your AI before anything goes live.
Brand store pages and stores analytics, built and measured by your AI.
Audience discovery and DSP audiences created, sized, and activated by your AI.
Off-Amazon attribution tags and performance, created and pulled by your AI.
Conversion events and campaign conversion tracking, configured and read by your AI.
v3 reporting, snapshots, and marketing mix modeling — requested and retrieved by your AI.
Brand metrics and benchmarks pulled by your AI to see where the brand really stands.
Tactical recommendations, partner opportunities, persona builder, and audience insights.
Call verified AMC administration, workflow, reporting, rule-audience, and ad-audience operations from an MCP-compatible AI client.
Real-time Amazon Marketing Stream subscriptions, wired up by your AI.
Reach forecasting and media planning, run by your AI before you commit budget.
Profiles, manager accounts, portfolios, billing, and account budgets managed by your AI.
Regions & markets
Pause an underperforming Sponsored Brands campaign across all 16 EU markets in one prompt. Pull AMC overlap from Japan and the US side by side. Switch markets without switching tools.
Every Amazon marketplace has its own quirks — different ad permissions, different DSP regions, different AMC instances. Kuudo handles the wiring so your chat just asks the right one.
Built in the open
We released the Amazon Ads MCP as open source — a contribution to everyone building on Amazon with AI, and working proof of what agent-grade enablement actually makes possible. There is no ceiling on where you take it. It is the same reason every one of our docs is public and our field guides walk you through it step by step: openness is how you take control, take ownership, and realize the promise we make in Why Kuudo.
We put it in the open for the whole Amazon-and-AI community, to show what's possible — not to fence anyone in.
Closed tools cap you at whatever a vendor decides to expose. In the open, the surface goes as far as your own ideas do.
Public docs, step-by-step guides, open code — the means to take control of the work, not rent it. You own what you build. Private. Trusted. Owned.
Coverage
Connect Kuudo as an MCP server inside the chat your team already uses. Ask for campaigns, listings, or reports in plain language. Answers come back grounded in live Amazon data.
Run multi-step goals against Amazon's APIs. Lean, efficient tools keep agents moving, so they finish the job instead of running out of room.
Drop Kuudo tools into the visual workflow builder you already operate. Trigger campaign updates, listing fixes, and report exports on a schedule or from any upstream event.
Wire Amazon data into the loop your engineers already use. Build dashboards, debug listings, and prototype agents without leaving the editor.
Guides
Real operator workflows, grounded in Amazon Agent Atlas, that run through this product.
Validate AMC SQL syntax, joins, and output shape in the sandbox before confirming real identifiers and runtime on a short production window.
Fix slow AMC SQL with early filters, smaller CTEs, safer joins, correct random sampling, and a staged timeout recovery test.
Every AMC query starts with four decisions: which table, which campaign identifier, which ad product type, and which date window. Get those...
AMC SQL rejects SELECT , ORDER BY and LIMIT, and drops rows describing fewer than 100 users. Here is what the agent checks before writing on...
Zero eligible DSP tactics is almost always a goal-KPI mismatch, not an outage. Create every entity paused, verify each write, then activate...
Measure web, app, store, and offline conversions in Amazon Marketing Cloud with Events Manager signals and a grounded total-impact SQL workf...
Find your optimal Amazon DSP frequency cap in AMC: bucket impressions and conversions, then cap where cumulative cost outruns cumulative ret...
Use AMC Campaign Groups to map DSP and sponsored ads paths into Sankey-ready source-to-destination rows you can chart without reshaping them...
Count new-to-brand customers in Amazon Marketing Cloud with the conversion-time table, without confusing users, purchases, or attribution dr...
Measure Sponsored Ads and DSP overlap in Amazon Marketing Cloud with aligned exposure groups, attribution timing, and campaign scope for a d...
Choose the right Amazon Marketing Cloud attribution source before custom modeling, audience building, or reconciling results with standard r...
Evaluate a Prime Day AMC lookalike against last year's rule-based audience using new-to-brand share, incremental reach, and cost per acquisi...
Build AMC CLTV segments with the four-segment taxonomy, the correct audience table variant, and a cohort migration diagnostic you can re-run...
Build high-value AMC audience seeds with the right table variant, a seed-size buffer, and the lookalike activation checks that keep them usa...
Build a lapsed-loyalist AMC audience with EXTEND TIME WINDOW, non-ad-exposed filters, and the activation-size checks DSP needs before launch...
Build a Prime Day AMC lookalike audience with seed SQL, sizing checks, and the 500 to 500,000 user window Amazon requires before activation.
Build a cart-abandoner audience in AMC with the right table variant, a seed-size check, and the DSP activation timing that avoids a stale se...
Measure Subscribe & Save lift in AMC with the repeatSnSOrder signal, the conversions all table, and a 90-day minimum window for a stable rea...
Package repeatable AMC analysis into agent Skills with Atlas retrieval, sufficiency checks, human approval, and run logging already built in...
Gate AMC audience activation behind human approval: size the seed, review the exact payload, then submit. 2000-user floor, 24 to 36 hour DSP...
Frequently asked questions
The bottom line
They raise the bid on the winner, cut the keyword that's burning spend, shift budget to what's converting, pause the losing ad, build the report — across every campaign type, around the clock. You set the limits and make the calls that matter. Brands grow ad sales without growing the team. Agencies manage more accounts without the all-nighters. Builders skip months of API plumbing.