Canonical page: https://www.kuudo.com/features/amazon-agent-flow/ Amazon Agent Flow # Agents that know your Amazon business — not guess at it. Your agents answer with real figures and take real actions on your Amazon business — across hundreds of accounts, from Claude, ChatGPT, or Cursor. It works because Kuudo lands your Selling Partner, Vendor, and Ads data in a private lake you own, then lets them act on it safely. [Explore the private beta](/contact/) [Read the docs](/docs/amazon-agent-flow/) 100s agency, brand & client accounts, one deployment Iceberg open tables across ads, orders & inventory 0 spreadsheets pasted into a chat /the agent lake ## Your Amazon data, in a private lake your agents can trust — on storage you own, ready to query from anywhere. Kuudo turns your Amazon seller, vendor, and ads data into an AI-native Iceberg data layer on storage you own — so private, trusted agents can safely query the data they need to drive your business from DuckDB, Snowflake, Databricks, or any Iceberg-aware engine. /table format ![Apache Iceberg](/assets/images/db/apache-iceberg.svg) ### Apache Iceberg The open table format every dataset lands in — schema evolution, snapshots, and time travel built in. Kuudo writes here /query engine ![DuckDB](/assets/images/db/DuckDB_icon-lightmode.svg) ![DuckDB](/assets/images/db/DuckDB_icon-darkmode.svg) ### DuckDB Query the lake in-process from a laptop or a function. No cluster to stand up, no warehouse to pay for. in-process SQL /object storage ![Cloudflare R2](/assets/images/db/cloudflare-color.svg) ### Cloudflare R2 Your own zero-egress bucket. Bring it; Kuudo provisions and writes to it. The keys never leave your account. zero egress /query engine ![Snowflake](/assets/images/db/snowflake-icon.svg) ### Snowflake Read the same Iceberg tables as Snowflake external tables — no copy, no reload, no second source of truth. external tables /lakehouse ### Databricks Query the same Iceberg tables through Unity Catalog and Databricks SQL — no copy, right beside the rest of your lakehouse. Unity Catalog /catalog + query ### AWS Glue / Athena Catalog the tables and run serverless SQL across the org — no servers to keep warm. serverless SQL /object storage ### Amazon S3 Land Iceberg tables right next to the rest of your AWS stack, in a bucket you already control. AWS-native Lake storage: Cloudflare R2, AWS S3 Tables or S3, and local Iceberg. External query engines connect through a compatible catalog and use their own access and cost controls. Owned by you Your bucket, your catalog, your keys. Kuudo provisions the lake and keeps it fresh, but the Amazon seller, vendor, and ads data lands in infrastructure you control — ready for your private AI agents, not locked inside someone else's platform. Cost-effective at agent volume DuckDB runs in-process for repeated agent queries. Agent Flow estimates the data a governed query would scan and enforces your scan ceiling before execution. Large results stay server-side for analysis, keeping raw tables out of chat context. Open, portable, no lock-in Iceberg and Parquet keep your data usable everywhere: DuckDB on a laptop, Athena in AWS, Snowflake, Databricks, or the next engine your team adopts. Your agents get a trusted data layer without tying your business to one query stack. Agent-ready by design Stable schemas, partition pruning, snapshot isolation, and access you control give AI agents the boring but critical foundation they need: reliable context, repeatable answers, and safe access to the Amazon data that drives your business. Amazon SP / Ads→ Kuudo workflows→ Parquet + Iceberg→ your R2 / S3 Tables / S3→ DuckDB · Snowflake · Athena · your agent /context for agents ## Agents need context for smart, efficient queries. We deliver it. Agent Flow tells your agents what data is available, what its fields contain, where it came from, and how fresh it is before they query. DuckDB runs the analysis, and compact results keep each follow-up focused on relevant evidence. ### Context before the first query Your agent gets dataset references, field definitions and classifications, freshness, and lineage before choosing what to query. It starts with information about the data already in hand, instead of spending queries trying to discover it. ### Room to test and follow up Agents work through questions in small steps: inspect, compare, check, refine. DuckDB runs in-process for that repeated analysis, and scan estimates let the agent size up a query before running it within your limits. ### Keep the reasoning focused Full results stay server-side for Python or curated analysis. The agent retrieves the relevant summary or small table for its next step, so its context stays focused on the evidence it needs to reason over. You set the scan ceiling for Agent Flow’s governed queries. Agents can narrow it for a task; execution enforces your limit. The query-limit FAQ below explains how this differs from a spending cap. MotherDuck’s analysis explains why context improves agent query accuracy and reduces the tokens needed to reach an answer. [Read “Agents Don’t Query Like Humans Do.”](https://motherduck.com/blog/agents-dont-query-like-humans-do/) [See the Agent Flow query and analysis workflow](/docs/amazon-agent-flow/) /taking actions ## Tell it what to do. It does the work — change a bid, fix a listing, ship an order. Your agent works inside your Amazon Seller, Vendor, and Ads accounts — not another read-only dashboard. You decide what each agent can touch, and every change it makes is logged. Counts below are API operations in Kuudo’s published Amazon MCP inventories. Agent Flow exposes its enabled operation catalog separately through discovery. /selling partner ### Sellers SP-API · 341 API operations · 3P Catalog products · ASINs * get catalog data * get ASIN matches Listings catalog · ASIN · variations * create a listing * update a listing * delete a listing Orders orders · shipments * get orders * confirm shipment Pricing & Fees Buy Box · repricing · fees * update price * update offers * get fees FBA Inbound & Inventory send to amazon · FBA * create inbound plan * update shipment * get inventory Multi-Channel Fulfillment off-Amazon · MCF * create fulfillment * cancel fulfillment External Fulfillment external · store · returns * update inventory * create shipment * confirm returns Amazon Warehousing AWD · distribution * create inbound * update shipment * get inventory Shipping & Buy Shipping rates · labels * get rates * create label * confirm shipment Feeds bulk · price · inventory * submit feed * get result Reports every report type * request report * get report Data Kiosk analytics queries * run query * get results Sales Analytics order metrics * get sales * get metrics Finances settlements · events * get settlements * get financial events Seller Wallet balances · payouts * create transfer * submit payout Tax & Invoices VAT · billing * submit invoice * create invoice A+ Content modules · brand story * create content * submit content Messaging & Notifications buyer · subscriptions * submit message * create subscription Customer Feedback reviews · feedback * submit review request * get feedback Subscribe & Save replenishment * create offer * update offer /vendor central ### Vendor VC-API · 1P Vendor Ops PO · ASN · 1P supply * update inventory * submit invoice * confirm shipments A+ Content modules · ASIN relations * update content * update ASIN relations * approve submissions /amazon ads + amc ### Ads Ads-API · 756 API operations · AMC Sponsored Products campaigns · keywords * create campaigns * update bids * create keywords Sponsored Brands v3 · v4 · creatives * create campaigns * update creatives Sponsored Display audiences · contextual * create campaigns * update targeting Sponsored TV streaming · video * create campaigns * update creatives Amazon DSP programmatic · line items * create campaigns * update line items Campaign Management unified · all programs * create campaigns * update budgets & bids Targeting keywords · products · negatives * create targets * update targets Creatives & Assets assets · moderation * create creatives * submit assets Brand Stores pages · analytics * update store * get analytics Audiences DSP · discovery * create audiences * get insights Amazon Attribution off-Amazon · tags * create tags * get attribution Conversions events · tracking * create conversion events * get conversions Reporting v3 · snapshots · MMM * request report * get report Brand Metrics metrics · benchmarks * get brand metrics * get benchmarks Recommendations insights · personas * get recommendations * get insights AMC workflows · audiences * create workflows * create audiences * run queries Marketing Stream real-time · subscriptions * create subscription * get stream Forecasts & Planning reach · media planning * get forecasts * get reach Accounts & Billing profiles · portfolios · billing * get profiles * create portfolio * get billing Fixes, not just flags Most tools tell you a bid slipped or a listing broke. The agent goes and changes it — then confirms the work is done. One ask, every account Run the same change across a single account or your whole book — agencies move every client at once, brands keep each marketplace in sync. You set the rules Decide what each agent can touch and what needs your sign-off. Every change is logged, so you always see exactly what it did. /what kuudo runs ## Get your Amazon data ready for agents — reliably, across every account. Amazon data was not designed for agents. It sits in proprietary tools, scattered exports, brittle reports, and workflows that still depend on people copying sensitive business data into chat windows — or giving AI broad access to personal computers. Kuudo changes the pattern. It pulls your Amazon seller, vendor, and ads data into one private place you control, then gives trusted AI agents safe ways to query, monitor, and act on it — without training the model to benefit your competitors. ### Agents do the work Kuudo is a native MCP server. Connect Claude or any MCP-compatible assistant and your agent can start an orders pull, request an ad report, check pipeline status, or query yesterday's numbers — all through approved actions and strict guardrails. Related [MCP overview](/features/mcp/) [AI clients](/features/ai-clients/) [Skills](/features/skills/) ### One connection to Amazon Selling Partner. Amazon Ads. Orders, listings, inventory, catalog, reports, campaigns, performance, Sponsored Products / Display / Brands, and Reporting v3. Discover the available operations and inspect their inputs before running or scheduling a report. Related [Amazon Ads MCP](/features/amazon-ads-mcp/) [SP-API MCP](/features/amazon-selling-partner-mcp/) [Vendor Central MCP](/features/amazon-vendor-central-mcp/) ### Pipelines that don't drop data Every run is durable. If Amazon throttles, a server restarts, or a job gets interrupted, Kuudo resumes from the exact step it left off — no duplicate loads, no silent gaps, no babysitting. Related [Cloud platforms](/features/cloud-platforms/) ### Built for hundreds of accounts One deployment can run hundreds of seller and ad accounts, hundreds of schedules, and account-specific pacing that respects Amazon's limits. Hourly here, daily there, backfill when needed — all config-driven. Related [Cloud platforms](/features/cloud-platforms/) [Skills](/features/skills/) ### You decide when fixes run The monitor ships observe-only. Review its audit log, then explicitly switch on automatic fixes. It can retry failed reads and recover pipeline work; it has no tools to change Amazon listings, prices, credentials, or accounts. Code enforces action caps, cooldowns, and attempt limits. Related [Docs](/docs/) ### Healthy checks use zero model calls Scheduled checks run for each tenant. Healthy cycles are recorded without calling a model; flagged anomalies go to AI triage. Every corrective action carries a recorded reason, so your team can review what happened and why. Related [Docs](/docs/) Durable · guardrailed · agent-ready ## Put AI to work on your Amazon business. Connect Amazon Selling Partner and Amazon Ads, and your agents can finally act on your data — safely, starting with your own accounts. Kuudo connects to Amazon Selling Partner and Amazon Ads. Capability availability may vary by connector and configuration. [Explore the private beta](/contact/) [Read the docs](/docs/) Frequently asked questions [Read MotherDuck’s agent-query analysis](https://motherduck.com/blog/agents-dont-query-like-humans-do/) ## Common questions. Is Agent Flow available through Managed MCP? + Agent Flow remains a Private Cloud capability. Its pipelines and query-ready lake run in your deployment and storage; Managed MCP covers Ads, Selling Partner, Vendor Central, Atlas and Iris. What is Amazon Agent Flow? + It lands your Selling Partner and Amazon Ads data in a private Apache Iceberg lake on storage you own, so agents answer with real figures from your business instead of guessing at them. Where does the data actually live? + In your own Cloudflare R2 or AWS storage, with AWS Glue Iceberg supporting S3 Tables and S3. A local Iceberg destination is also available. Cloud deployment options and implemented lake storage backends are separate choices; Google Cloud Storage and Azure Blob are not currently Agent Flow lake backends. What can query the lake? + Anything Iceberg-aware. DuckDB in-process from a laptop or a function, Snowflake as external tables, Databricks, or your own engine, reading the same tables with no copy and no second source of truth. Why land the data at all instead of calling the Amazon APIs live? + Because agents need to answer across whole accounts and long time ranges, which is slow and rate-limited against the APIs directly. The lake makes those questions cheap to ask and keeps history that the APIs age out. Why use DuckDB for agent queries? + Agents often inspect schemas and ask several small follow-up questions before reaching an answer. DuckDB runs in-process, which suits that pattern without a separate warehouse cluster. MotherDuck reported 29 times as many agent queries as human queries in its September 2026 analysis, alongside frequent schema inspection and short gaps between queries. Those figures describe MotherDuck’s workload. Agent Flow combines DuckDB with dataset context, scan limits, and server-side analysis. Can an agent raise its own query limits? + No. For Agent Flow’s governed queries, you set the server’s scan ceiling. An agent can request a lower limit, but cannot raise yours. The estimate measures data bytes, not a dollar price or a cap on your total cloud bill. Queries made through a separate engine use that engine’s controls. Does the monitor fix problems automatically? + It ships observe-only: it watches and reports without taking corrective actions. After reviewing the audit log, you can explicitly enable automatic remediation. The monitor can retry failed reads and recover pipeline work within action caps, cooldowns, and attempt limits. It has no tools to change Amazon listings, prices, credentials, or accounts. Can an agency unify Seller Central and Amazon Ads data across clients? + Yes. Agent Flow lands each client's Selling Partner and Amazon Ads data in a private lake, and hundreds of brand or client accounts run from a single deployment, across datasets covering ads, orders, and inventory. Your AI then answers one question across both, such as ad spend next to inventory, instead of joining exports by hand. Do we need a data warehouse first? + No. DuckDB queries the lake in-process, so there is no cluster to stand up before your agents can use the data.