Kuudo runs one MCP server per Amazon surface. Each server exposes domain-scoped tools so agents can read, analyze, and safely queue changes — this page is the index: what's available today, and where to find the full tool-by-tool reference for each one.
MCP servers
Amazon Ads MCP
Sponsored Products, Brands, Display, and TV, plus DSP (demand-side platform), AMC (Amazon Marketing Cloud), the unified Ads API v1, attribution, and reporting — see the Ads MCP tool reference for the live resource and tool counts.
See every Amazon Ads MCP tool ->
Amazon Selling Partner MCP
Catalog, listings, orders, pricing, FBA (Fulfillment by Amazon), fulfillment, finances, and more across the seller side of the Selling Partner API — see the SP MCP tool reference for the live resource and tool counts.
Repository and registry entry: github.com/KuudoAI/amazon-selling-partner-mcp · io.github.KuudoAI/amazon-selling-partner-mcp
See every Amazon Selling Partner MCP tool ->
Amazon Vendor Central MCP
Purchase orders and acknowledgements, direct fulfillment, shipments and invoices, and transaction status across the vendor side of the Selling Partner API, with Vendor Retail Analytics, Data Kiosk, and A+ Content reached through the shared Selling Partner resources — see the Amazon Vendor Central MCP page for the live resource and tool counts.
Repository and registry entry: github.com/KuudoAI/amazon-vendor-central-mcp · io.github.KuudoAI/amazon-vendor-central-mcp
See the Amazon Vendor Central MCP ->
Amazon Agent Iris
Creates and edits Amazon listing images and product video from live catalog data, with the argument schemas published in the catalog — see the Amazon Agent Iris page for what it can do.
Repository and registry entry: github.com/KuudoAI/amazon-agent-iris-mcp · io.github.KuudoAI/amazon-agent-iris-mcp
Amazon Agent Atlas
A retrieval layer your agent checks before it answers, indexing Amazon operating knowledge across Ads, Seller Central, and Vendor Central — see the Amazon Agent Atlas page for the live corpus size.
Repository and registry entry: github.com/KuudoAI/amazon-agent-atlas-mcp · io.github.KuudoAI/amazon-agent-atlas-mcp
More servers are on the way.
What connecting costs
When an MCP client connects, the server hands it the definitions of every tool it advertises, and those definitions occupy the context window before any tool is called. The Amazon Ads MCP advertises 3 tools totaling 388 tokens at connect, 0.19% of a 200k window, and indexes the Amazon Ads API operations behind them. Adding an operation to the catalog adds no tokens at connect.
Amazon's official Amazon Ads MCP Server advertises 110 tools totaling 92,678 tokens at connect, 46.34% of a 200k window. Both servers were measured on 2026-09-12 with the cl100k_base tokenizer; the method and the open items are in Amazon Ads MCP Token Cost: Compare Tool Catalogs.
Rate limits, throttling and long-running jobs
Each Amazon endpoint behaves differently, and Amazon decides the pace: its rate limits, how long a report or export takes, and whether a job completes or fails. Kuudo's MCP servers pass Amazon's current state to your AI with every call:
- The rate limit and how much of it remains, from Amazon's response headers.
- When to retry after Amazon throttles a request, and whether an error is worth retrying at all.
- How long a report or export usually takes, the statuses it moves through, and which tool to call next.
Your AI decides when to wait, retry or move on. For long-running and scheduled work, Amazon Agent Flow runs the job itself: it schedules each status check, reuses an equivalent request instead of duplicating it, and retries read-only runs after Amazon throttles, never runs that change Amazon data.
Read vs. write tools
Every tool on every server is tagged read or write. Read tools return scoped data and are safe for normal agent use — exploration, diagnostics, reporting. Write tools are guarded: they validate input, apply workspace policy, and may require explicit human approval before changes reach Amazon.
Auditability
Every tool call records the requesting user, agent, workspace, parameters, result summary, and approval state, across every server. Use the run log to reproduce outputs or review changes before activation.