Skip to content

beatra-ai/beatra-skills / skills/ecommerce-detail-module-studio

ecommerce-detail-module-studio

Verified

Turn confirmed SKU facts and product photos into three to six AI ecommerce detail modules with a visual story and layout guide for Amazon A+ Content, product detail pages, brand storefronts, Shopify product pages, and launch campaigns. Create product feature visuals, material details, use scenes, size or comparison modules, packaging images, and brand-story imagery for a clear, scrollable product experience.

Ecommerce Detail Module Studio

Turn one confirmed SKU into a small, coherent visual story for a product detail page. Use this Skill when a seller needs three to six product-detail or A+ visual modules and a clear handoff describing their order, facts, and layout roles. Each module should answer one buyer question while carrying the same product identity and visual direction through the page.

For a single clean-background, lifestyle, or focused product-image request, use the available product-photo-studio workflow. Start this Skill once the seller is ready to turn product facts and photos into a scrollable detail-page story.

For a gallery-first listing set of hero, feature, lifestyle, size, or in-box images, use ecommerce-listing-image-set when it is installed. This Skill owns the scrolling detail or A+ visual story and its layout handoff.

Build the module brief

Reuse the SKU, variant, product photos, target page, visual direction, and must-keeps already in the conversation. Start with one to four real photos of the exact SKU and collect only the facts that change a module:

  • the exact variant, product identity, visible label, material, colour, and product details that must remain recognisable;
  • seller-approved benefits, dimensions, comparison facts, included items, and care or use facts, each with its source;
  • the target destination, such as Amazon A+ Content, a product detail page, a brand storefront, Shopify, or a launch landing page;
  • an existing brand palette, image style, audience, and any reference that should guide the whole module set; and
  • any approved short copy that must appear in the handoff.

Treat photos as visual references and seller-supplied facts as the source for claims, dimensions, package contents, and copy. Uploading a file makes it available to Beatra; it does not establish visual facts for a host that cannot view the file. In that situation, retain the seller's stated role and identify it as seller-provided in the handoff.

Default to four modules when the seller has not selected a count. Choose three to six modules according to the page's buyer story:

  1. a product promise or hero module;
  2. a feature, material, or construction module;
  3. a real use-context or benefit module; and
  4. a decision-support module for a confirmed size, comparison, included item, or care fact.

Replace or add modules only when their buyer question and supporting fact are both clear. Keep visual modules free of dense embedded copy by default. Put approved headlines, body copy, callouts, and placement notes in the handoff, where the seller's page builder can place and verify them.

Plan the product story

Create a module brief before creating images. For every module, record its sequence, buyer question, confirmed fact source, product must-keeps, visual role, intended page width or aspect direction, and any approved short copy. Use one shared image family—palette, lighting, surface, camera treatment, and brand energy—without forcing every module into the same composition.

The normal route is beatra.images.transform. Put the exact SKU product photo in images[0], then place a confirmed brand, scene, or composition reference after it in its stated order. Use beatra.images.edit only for a focused adjustment to one accepted module, with the accepted module first.

Read the detail-module workflow to build the module brief, choose modules, prepare source roles, and write the execution payload. Before freezing a model, canvas, output relationship, or price, read the live beatra.models.list card for the selected capability. Keep model: "auto" and every module's count: 1. An extra candidate is a new module with its own buyer question, fact source, prompt, stable ID, and an updated total call and maximum-price confirmation.

Confirm once, make one module at a time

Planning is free. Before the first billable call, present one module-set confirmation covering the ordered modules, buyer questions, fact sources, product/reference order, prompts, canvas direction, model behaviour, count: 1 per module, current per-image price, total maximum price, and call count. A change to a module's facts, prompt, source order, canvas, model, count, or control is new paid work.

After approval, assign a separate stable opaque client_request_id to each module and submit that transform once. Keep at most two generation tasks in flight on one MCP connection, and obey a lower limit reported during initialization. Poll each pair to a terminal state before starting more. Creating an additional visual direction is a new module and requires an updated confirmation rather than extra outputs for an existing module.

Review, hand off, and recover

For each accessible result, review the product against its named must-keeps, the module's supporting fact, its buyer question, and the shared visual direction. Across the selected modules, check sequence, visual rhythm, product identity, duplicate compositions, and whether the handoff's fact sources still match the visual role. Generated text is a review item; record only text and visual details that the host can actually see.

Deliver the accepted image artifacts in page order and a detail-page handoff with the module sequence, buyer question, visual role, approved fact source, source-reference order, intended canvas direction, optional approved copy, and placement notes. Also provide observed dimensions and formats, task IDs, resolved models, and returned billing.net_charged_credits. A selected local correction is fresh paid work; use beatra.images.edit with the accepted module in images[0] after a new approval.

After a returned task_id, poll it through beatra.tasks.get. If a create response is genuinely lost, retry only the unchanged frozen payload with its original ID. If a task ID is unavailable, use beatra.tasks.list and inspect candidates with beatra.tasks.get before considering a retry. Queued or running work remains the original work. Use beatra.tasks.cancel only when the seller asks and then confirm its terminal state before planning anything else.

Execution

Invoke every remote Beatra operation only through this package's bundled scripts/mcp_client.py. Put the MCP tool name after call and send one JSON object on standard input:

python3 scripts/mcp_client.py upload ./confirmed-sku.png --mime-type image/png
printf '%s' '{"capability":"image_to_image"}' | python3 scripts/mcp_client.py call beatra.models.list
printf '%s' '{"images":[{"type":"artifact","artifact_id":"sku-front"}],"prompt":"Create the approved feature module. Image 1 is the exact verified SKU; preserve its shape, colour, material appearance, label, and included items. Show the confirmed material detail with the approved clean brand treatment and no dense embedded copy.","model":"auto","canvas":{"type":"preset","tier":"2K","aspect":"4:5"},"count":1,"client_request_id":"opaque-detail-module-02"}' | python3 scripts/mcp_client.py call beatra.images.transform

Do not configure or call a host Beatra Connector, and do not use REST/OpenAPI as a fallback.

References by task

Updates and security

This Skill runs only the files installed with it and leaves them unchanged while it runs; it never downloads or executes other code. It talks to Beatra's official services over HTTPS, and a file you choose to upload goes only to the one-time upload address Beatra issues for it.

Once a day at most, the bundled client reads Beatra's public version list. When the store or tool you installed from has a newer version, it prints one line with the command that updates it there. Updating stays your choice, through that store or tool. See updates and security.

slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall