Check product photos against marketplace image rules before listing
Before a new listing goes live, a decision model checks the main product photo against the marketplace's image rules and holds back any photo that seems to break one, so a person can fix it first.
Uses image input, which TypeSafe AI labels a preview feature for Jev. Test it on your own sample images before relying on it.
An example pattern written by Agents Lookup. We have not yet found a public build of this exact job.
The use case
Marketplaces reject or suppress listings when the main photo breaks their rules, such as a background that is not plain white, added text or watermarks, or a product that fills too little of the frame. Checking every photo by eye is slow when a catalogue grows. Jev looks at each main image and answers a short set of yes-or-no questions, one per rule you care about, each with a probability. A photo passes only when every rule clears your threshold, for example 0.9. Any photo that fails a rule, or that Jev is unsure about, is held with the failed rule named. A person still decides whether to reshoot, edit, or list anyway, and the marketplace's own review has the final word.
Example workflow
- When a product is created or updated in Shopify, a Make scenario collects the main image URL and SKU.
- The image is sent to Jev with one Noul question per rule: the background is plain white, the image has no added text or watermark, the product fills most of the frame, and only the product for sale is shown.
- Code compares each probability with the threshold you set and marks the photo pass, fail, or unsure.
- Passing photos are tagged ready in a Google Sheet so the listing can be pushed to the marketplace on the normal schedule.
- Failed or unsure photos are posted to Slack with the rule that tripped, so whoever handles images can edit or reshoot.
- Once a month, compare Jev's holds with any marketplace rejections and adjust the rules or threshold.
Tools
- Jev (TypeSafe AI)
- Shopify
- Make
- Google Sheets
- Slack
- Canva
Risks and guardrails
- Image input is a preview feature in Jev 1.14 and TypeSafe says it is not production-ready yet — guardrail: test it on a few hundred of your own photos against a person's labels before relying on it, and keep a person checking every answer until the results hold up.
- Marketplace rules change and differ by category — guardrail: keep the rule questions in a sheet you can edit, and check them against the marketplace's current image guide each quarter.
- A light grey studio background may be judged white or not white inconsistently — guardrail: set a higher threshold for the background rule and send borderline photos to a person.
- A passed photo can still be rejected by the marketplace — guardrail: Jev only decides what to hold back; it does not promise acceptance, and rejections are logged for review.
Related recipe
Shopify Order-Exception Triage Agent: An agent reviews new Shopify orders for address problems, risk flags, stock gaps, split shipments, and customer notes, then sorts them, routes them, and drafts customer messages for a person to approve.
Want it set up for you? Request a quote.
Sources
Written by Agents Lookup. Credits and official pages for the tools this use case names:
- Shopify Dev MCP server (AI Toolkit) (official page): shopify.dev/docs/apps/build/ai-toolkit (checked 2026-11-09)
- Make MCP server (official page): developers.make.com/mcp-server (checked 2026-11-09)
- Slack MCP server (official page): docs.slack.dev/ai/slack-mcp-server (checked 2026-11-09)
- Canva Skills (official page): github.com/canva-sdks/canva-claude-skills (checked 2026-11-09)