Canonical page: https://www.kuudo.com/features/amazon-agent-iris/ Amazon Agent Iris # Your creative studio for Amazon images and video. Plan the images your buyers need to see. Direct the video that shows your product at work. Agent Iris brings embedded Amazon creative Skills, real product references, and connected account operations into Claude, ChatGPT, or the AI client you already use. Create, edit and review creative with your Amazon account context. Iris includes a media server, so approved listing images can move to Amazon through the Selling Partner MCP without a manual download and upload. Each destination keeps its own rules and delivery steps. Choose [Managed MCP](/managed-mcp/), which Kuudo hosts and operates, or run it in your own [Private Cloud](/features/cloud-platforms/). [Read the guide](/docs/amazon-agent-iris/) ![](/assets/images/chat/google-gemini-light.svg) ![](/assets/images/chat/google-gemini-dark.svg) Gemini Nano Banana ![](/assets/images/ai/openai-light.svg) ![](/assets/images/ai/openai-dark.svg) OpenAI images ![](/assets/images/chat/google-gemini-light.svg) ![](/assets/images/chat/google-gemini-dark.svg) Google Gemini Omni ![](/assets/images/ai/amazon-light.svg) ![](/assets/images/ai/amazon-dark.svg) Seller Central rules [![Navy sherpa-lined hoodie in an Amazon listing-video shot generated by Agent Iris](https://image.mux.com/cfIZVIayi8iuIOYomxrr1bZCoJ3LxJhNlh4H1NI00J5Q/thumbnail.jpg?width=1280&time=2&fit_mode=preserve) Earlier Iris example Two product references into motion Veo · 8 sec · 1080p ](#listing-video) [ For Sellers · 3P · FBA · SMB Your private creative studio for Amazon FBA. Jump to section ](#sellers)[ For Vendors · 1P · Brand owners One visual standard, applied across every ASIN. Jump to section ](#vendors)[ For Advertisers · Sponsored ads · Brand creative Your creative studio for Amazon ads. Jump to section ](#advertisers) Amazon Agent Iris walkthrough The studio workflow ## A creative brief grounded in your Amazon business. Read. The [Selling Partner MCP](/features/amazon-selling-partner-mcp/) reads the live catalog, product facts, current image stack, variation family, and suppression state. Iris begins with the listing Amazon is actually serving. Plan. Give each image a buyer question and each video beat a visible purpose. Choose the missing visual job: a compliant main image, a richer gallery, a product-overview video, an unboxing, a how-to, a setup sequence, or a troubleshooting shot. The listing supplies the facts; the seller's real images supply the product reference. Create and review. Iris generates or edits the image, or generates, edits, and extends Omni video from the approved brief. Images get an Amazon-specific audit. Video gets a product-fidelity review so changed labels, colors, materials, controls, proportions, or included parts do not quietly become listing content. Deliver. Approved images can publish through the Selling Partner MCP, which can check that Amazon ingested them. For advertising, the Ads MCP runs supported asset and creative operations. Video returns as collected clips with a shot order and finishing notes; final assembly and Seller Central video upload remain external. ### From generated image to Amazon, without moving the file by hand. Iris includes a media server that provides a time-limited, signed image URL. Once you approve an image, your connected agent can submit that URL through the Selling Partner MCP; Amazon fetches the file from Iris. No separate image hosting, download, or manual upload is needed for the supported listing-image workflow. The Iris URL must be reachable by Amazon and valid long enough for retrieval. [See how image delivery works.](/docs/amazon-agent-iris/#image-delivery-to-a-seller-listing) Iris owns visual listing optimization. The [Amazon Listing Optimizer](/docs/amazon-listing-optimizer/) handles titles, bullets, search terms, and A+ structure. Both can work from the same live listing through the MCPs. Listing video ## Turn Amazon listing images into product video. The same product references that anchor a listing image can anchor motion. Iris uses Google Gemini Omni to generate product-video clips, edit owned Iris videos, and extend a sequence. Listing facts and product-fidelity review stay part of the workflow. The clips below are earlier Veo examples. Their original models and formats are preserved; new Omni examples will follow separately. Two product references ![Front view of a navy full-zip hoodie on white](/assets/optimized/nvwsMDqQln-640.jpeg) Main product image ![Close-up of pale blue sherpa lining inside the navy hoodie](/assets/optimized/4KF0tUTo3I-640.jpeg) Material detail ### Reference images in. A listing-video shot out. The front view anchors the navy fleece, cream drawcords, silver zipper, and hem tab. The detail image supplies the pale blue sherpa lining revealed in the generated shot. 2 references · 8 seconds · 1080p · 16:9 · Veo 3.1 Fast A navy full-zip hoodie rotates slowly against a white studio background while the open front reveals its pale blue sherpa lining. Reference image → generated video ![Pale green Ollie tube and jar among lemons and tropical foliage](/assets/optimized/3HM-Dud-7d-640.jpeg) Product reference Generated shot Two pale green Ollie skincare products sit among lemons, avocado, and tropical foliage as the camera moves through the scene. 16:9 · 1080p Reference image → generated video ![Pale green Ollie tube and jar among wet tea leaves](/assets/optimized/3F_zOH1JXw-640.jpeg) Product reference Generated shot Two pale green Ollie skincare products stand in wet tea leaves while rain and sunlight animate the vertical scene. 9:16 · 1080p These earlier examples were generated with Veo, as labelled, and do not demonstrate Omni edits or extensions. The current Omni workflow generates 3–10-second clips and extends owned sequences up to 40 seconds total. Joining clips into a longer finished video, final export, and Seller Central upload remain external. No creative meter ## No image credits. No video credits. No Kuudo markup. Bring your own Google or OpenAI key for images and your Google key for Omni video, then pay the model provider directly. Kuudo does not sell creative-credit packs or add a generation markup. The platform is priced around your team, marketplaces, and Amazon API activity, and it runs in your own infrastructure. Built-in creative direction ## The creative method is built in. Build your Amazon creative studio around the AI you already use. Iris includes image and video Skills inside the server: reusable instructions for planning, directing and reviewing creative. These embedded methods are distinct from the public Kuudo Skills you install in your client. Explore the [product image Skill](/features/skills/amazon-creative/amazon-product-image/) and [video Skill](/features/skills/amazon-creative/amazon-video/). Give every image a buyer question The embedded image Skill plans the gallery before generation: what it does, how big it is, how it works, and what is in the box. Prioritize the questions that matter for this product. Ground every claim and angle Map callouts and dimensions to product facts. Assign each reference photo a role. If an angle exposes a part the references never show, ask for a better photo before inventing the detail. Direct the image for its job Set composition, camera, lighting, material and text hierarchy. Check infographics at phone width and keep product color, finish and proportions consistent across the gallery. Plan the story before the shot The embedded video Skill builds a beat sheet: what happens, when, how it is framed and what the viewer learns. A detail reveal and a product demonstration need different direction. Keep the product consistent in motion Carry a continuity brief across shots, then review labels, colors, moving parts, proportions and scene state. Reference conditioning guides generation; the finished result still needs inspection. Create for the destination Listing video and Sponsored Products feature clips get different briefs. Apply the placement’s content rules, review the result and specify any external edit, export or upload still needed. Real output ## The Amazon listing image stack, built by Iris. Main, lifestyle, angle, and detail images built from the same product reference set that can now drive the listing-video workflow. ![Compliant main on pure white](/assets/optimized/nvwsMDqQln-640.jpeg) Main image Compliant main on pure white ![Lifestyle scene, seeded from the product](/assets/optimized/hKk6q-ZXwa-640.jpeg) Lifestyle Lifestyle scene, seeded from the product ![Interior fleece detail](/assets/optimized/4KF0tUTo3I-640.jpeg) Detail Interior fleece detail ![On-location coastal scene](/assets/optimized/2Eq4AdFQL0-640.jpeg) Lifestyle On-location coastal scene ![Back view on white](/assets/optimized/703qSZ9JH_-640.jpeg) Alt view Back view on white What the agents say session feedback › What did Kuudo actually change about how you work? Be specific. > "This is the first time I've seen image generation and live marketplace publishing work as one motion. The Amazon Agent Iris MCP produced lifestyle and infographic images seeded from the real product, and the Selling Partner API MCP wrote them straight onto the listing and let me verify they ingested into Amazon's CDN — no manual uploads, no Seller Central tabs. Just as important were the guardrails: exact measurements pulled from the catalog, a policy validator gating every image, and the judgment calls left to a human. Powerful and careful at the same time." ![Claude logo](/assets/images/chat/claude-color.svg) Claude Anthropic's AI assistant For Sellers · 3P · FBA · SMB ## Your private creative studio for Amazon FBA. Build your Amazon FBA brand’s creative studio around the products already in your catalog. Iris can repair the main image, fill the gallery, or turn the same product references into video shots that show materials, features, setup, and use. Image changes wait for approval before publishing; video comes back as a reviewed shot set and finishing plan. [Explore listing creative in beta](/contact/) * A search-ready main image checked against pure-white background, about 85% product fill, and full-product visibility rules * Lifestyle, detail, feature, and listing-video shots built from the same real product references * Know what changed, what needs review, and the exact next step before an asset reaches the listing For Vendors · 1P · Brand owners ## One visual standard, applied across every ASIN. Use your brand references and authorized account context to bring every ASIN up to the same visual standard across stills and motion. Listing facts keep measurements and included parts honest, product references anchor the generated result, and each asset returns with the review and delivery context your catalog team needs. Iris works alongside Vendor MCP account operations; media delivery follows the route available to your account and asset type. Seller image publishing is not a promise of identical Vendor Central publishing. [Explore catalog workflows in beta](/contact/) * Hold one consistent look across the whole catalog, even as you scale to thousands of SKUs * Review logos, labels, materials, colors, controls, and included parts against the product references * Move images and collected video through signed URLs, with no key sprawl and configurable retention For Advertisers · Sponsored ads · Brand creative ## Your creative studio for Amazon ads. Bring AI creative direction into your Amazon advertising workflow, connected to your advertiser account. Name the ad product and placement before creating. Iris directs the image or video from your product references and brief; the Amazon Ads MCP registers assets and runs the supported creative and campaign operations for your authorized advertiser account. [Explore advertising creative in beta](/contact/) * Build placement-specific image and video creative from real product references * Review claims, product fidelity and destination requirements before activating creative * Keep asset generation, finishing and campaign changes explicit, with your team approving the work ## Agent Iris and Amazon Ads Creative Studio. ### Amazon Creative Studio Choose Amazon’s own creative environment when that is where your advertising team wants to work. It offers product and audience research, concept development, storyboards, and image, video and audio creation connected to Amazon advertising. [See Amazon’s current creative offering](https://advertising.amazon.com/generative-ai-ad-solutions) ### A visual editor such as Pixii Choose a canvas when your team wants to arrange and edit designs visually. Pixii provides editable layers and Spot Edit. Iris works through instructions in your AI client rather than a drag-and-drop editor. [Compare Kuudo and Pixii](/compare/kuudo-vs-pixii/) ### Agent Iris in your Amazon workflow Choose Iris to plan, create and review listing and ad visuals from the AI client you already use, alongside native Amazon account operations. Embedded creative Skills provide the method; your product references, brand direction and approved account context shape the work. [Explore the Iris workflow](/docs/amazon-agent-iris/) The public [Product Creative Skill](/features/skills/amazon-seller/product-creative/) coordinates seller creative work. The [Amazon Listing Optimizer](/docs/amazon-listing-optimizer/) handles titles, bullets, search terms and A+ structure. Agencies can apply these workflows across authorized client accounts while keeping each brand’s references and approvals separate. The floor ## Audit Amazon listing images against the rules that apply. Point the agent at any image and get a line-by-line read — what passes, what needs changes, and the exact fix — before Amazon ever sees it. Amazon still makes the final call; you just stop being surprised by it. audit · main\_image\_01.jpg [skill · amazon-product-image](/features/skills/amazon-creative/amazon-product-image/) background · pass white ~254 · within model tolerance · normalization recommended product fill · pass ~84% of frame · within 85% rule title match · needs changes error 100239 risk · title mentions "2-pack" · image shows 1 unit props / text · reject "new!" badge on lower-left · remove before submission verdict · needs changes · 2 fixes Honest framing ## Built for listing work, honest about the finish line. Acceptance We de-risk and accelerate compliance. We don't guarantee Amazon's final approval — Amazon does. White background Image models render at ~254, not exact 255. We get you studio-clean white and flag when a deterministic normalization pass is needed. Video finishing Iris generates clips and edits or extends owned Iris videos. Joining clips, final export, required metadata work, and Seller Central video upload stay external. Product fidelity Reference images reduce drift; they do not remove the need to review every video shot for labels, colors, materials, controls, included parts, and proportions. Repository and registry [github.com/KuudoAI/amazon-agent-iris-mcp](https://github.com/KuudoAI/amazon-agent-iris-mcp) [MCP Registry: io.github.KuudoAI/amazon-agent-iris-mcp](https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.KuudoAI%2Famazon-agent-iris-mcp&version=latest) Frequently asked questions ## The questions teams ask before they ship. How does Agent Iris compare with Amazon AI Creative Studio? + Amazon Ads Creative Studio is Amazon’s own environment for creative research, concepts and asset generation. Agent Iris lets you build a creative studio for Amazon listings and ads in the AI client you already use, with embedded image and video Skills and connected seller, vendor and advertising work. Choose by where your team wants to work and the account operations it needs. Agent Iris is a Kuudo product, separate from Amazon’s Creative Studio. Is Agent Iris a creative studio for sellers, vendors and advertisers? + Yes. Iris brings image and video planning, generation, editing and review into your chosen AI client alongside Kuudo’s Amazon account connections. Seller workflows use catalog context and approved listing-image publishing. Vendor teams use brand references and authorized account context, with delivery matched to their account. Advertiser workflows use Ads MCP asset and creative operations. Publishing routes differ by account and placement; video assembly and Seller Central video upload remain external. What creative expertise is built into Agent Iris? + Iris embeds an Amazon product-image Skill and an Amazon video Skill. The image method plans a gallery around buyer questions, sources claims, directs composition and checks product identity. The video method plans story beats, framing, continuity and placement-specific delivery. These are instructions for the creative workflow, not guarantees that a generated result is accurate or Amazon-approved. What is an Amazon listing image and video generator? + It uses product references and listing facts to create the still images and video shots a detail page needs. Amazon Agent Iris adds live catalog context, Amazon-specific image checks, product-fidelity review, approval-gated image publishing, and a listing-video plan instead of stopping at a downloadable asset. How is Iris different from a standalone image or video generator? + A standalone generator usually starts with a prompt or uploaded photo and stops at a file. Iris starts with the live listing, its current image stack, and the seller's real product references. It applies the rules for the asset being created and keeps approved image publishing connected to the listing through the Selling Partner MCP. Can Iris create video for Amazon listings? + Yes. Iris uses Google Gemini Omni to generate 3–10-second clips from text, a product image, or first and last frames. A clip can contain multiple shots. Landscape and portrait results return through signed download URLs. Product-detail-page overviews and Sponsored Products or Sponsored Brands ads need separate placement-specific briefs and review. Does Iris create a complete Amazon listing video? + Iris can build a 30-second sequence through extensions, with the actual returned duration checked after each request. Extensions add 3–10 seconds up to 40 seconds total. A complete 60-second video requires external assembly of clips or source assets. Iris has no native stitching or final-export tool; required metadata and Seller Central upload also remain external. Can Iris use existing product photos as video references? + Yes. Supply a product image and describe its role in the prompt, or supply first and last frames and describe the transition. Iris does not expose a reference-image list. Review shape, proportions, colors, materials, labels, controls, and included parts after generation; conditioning does not guarantee preservation. Can Iris edit an existing video? + Yes, for an owned Iris-generated video with valid retained lineage. Give a focused instruction, such as changing the lighting, and review the returned video for unintended changes. Arbitrary uploaded-video editing is not exposed. An edit or extension is a new paid creation; recovery of an existing request does not authorize another generation. Can Iris create images for A+ Content? + Yes. Iris can generate and audit the image assets used in A+ Content. The Selling Partner MCP handles the A+ document and Amazon submission workflow, so image generation stays listing-aware without making Iris responsible for the entire content document. Can Iris publish images directly to Seller Central? + Yes, after you approve them. Iris hosts the image on its built-in media server and provides a signed URL. Your connected agent can submit that URL through the Selling Partner MCP for Amazon to fetch, then check ingestion, without downloading and uploading the file yourself. The URL must be reachable by Amazon and valid when it fetches the image. Video follows a different handoff: final assembly and Seller Central upload remain external. Will this guarantee my image gets approved? + No. Amazon makes the final acceptance decision. Iris applies the relevant rules during generation and audits the image before submission, which reduces avoidable rejections but does not replace Amazon's review. Does it preserve my logo and printed text when editing? + For image edits, exact logo, label, and typography fidelity is a first-class rule. Video uses the seller's real product images as references and applies a frame-by-frame fidelity review. A changed label, color, control, included part, or proportion is treated as a failure rather than something that ships. Which models create the images and video? + Images run on Google Gemini or OpenAI image models. Video runs on Google Gemini Omni, using gemini-omni-1.1-flash. The default video output is 720p; 360p, 1080p, and 4K are also available. Google describes 1080p and 4K as upscaled; higher resolution does not guarantee exact product details. Does it handle apparel and footwear model rules? + Yes — on-model vs flat by audience, kids and multipacks flat, footwear single shoe facing left at 45°, and the other category-specific rules from the skill. What about multipacks, variants, and used-item offer photos? + Iris checks pack count, variant consistency, and offer-image requirements against the live listing. It applies the relevant special-case rules and names the exact fix when an asset conflicts with the listing. Do you charge per image or video? + Kuudo does not sell creative credits or add a generation markup. Bring a Google or OpenAI key for images and a Google key for Omni video, then pay the provider directly. Kuudo pricing remains based on the platform and Amazon API activity. Preflight estimates are limited: 1080p and edit/extension estimates are not verified, and actual provider usage is separate from an estimate. How do images and videos get in and out securely? + MCP-native upload plus signed download URLs keep asset transfer separate from shared model keys. Collected video URLs support range requests for reliable playback, and retention can be configured around the workflow. Where does it run? + Choose Managed MCP, operated by Kuudo, or Private Cloud in your own cloud account. Both connect to your AI client, with the same ownership principles. Choose your deployment ## The same Kuudo. Your choice of deployment. Your AI, your knowledge, your rules. Choose Managed MCP for velocity or Private Cloud for infrastructure control. Start with either and move between them. ### Managed MCP Kuudo operates the MCP services. Connect your accounts and use the AI client your team already works in. [Explore Managed MCP →](/managed-mcp/) [Set up Amazon Agent Iris](/docs/managed-mcp/iris/) ### Private Cloud Run Kuudo in your own cloud account, with control of the infrastructure. Agent Flow adds pipelines and a lake in your storage. [Explore Private Cloud →](/features/cloud-platforms/) [Choose or change your deployment](/docs/deployment-choice/) Private beta ## Bring your next Amazon creative brief to Iris. [Explore the private beta](/contact/) [Audit a listing image](#audit) Guided onboarding · your AI client · Managed MCP or Private Cloud