Makes on-model clothing photos for online stores: pick a model and scene, shoot, approve frames.
Scored across 14 tools
Most tools target clearly different resources and actions (products, scenes, shoots, frames, account), and the descriptions explicitly cross-reference related tools to steer selection. The main soft spot is the trio list_models / list_styles / list_scenes, which all serve 'choosing a scene' and could momentarily be confused, though each lists a distinct entity type.
Every tool follows a strict snake_case verb_noun pattern: add_*, get_*, list_*, approve_frame, remove_frame, request_scene, start_shoot. No camelCase, no vague single-word verbs, and the verb choice (get vs list vs add) is used consistently across resource types.
14 tools sits squarely in the well-scoped range, and each maps to a distinct step in the product-photography workflow (account, catalog, scene selection, shoot start, polling, approval). Nothing appears redundant or gratuitous.
The core lifecycle is covered end to end: add product and photos, browse models/styles/scenes, request a scene, start a shoot, poll results, list frames, approve and unapprove. Minor gaps exist (no product deletion/update or image removal, no way to cancel a queued shoot), but these are partly delegated to the host store platform and rarely block an agent.
Adds a product to the store's Shootika catalog, optionally with photos fetched from public image URLs. Use it when a product to shoot is not in list_products yet. The product is created even when a photo cannot be fetched; those come back in images_failed. Not available on Shopify stores: their products come from Shopify, so add the product there instead.
Inputs: title (required), family, image_urls, external_id, +2 more.
Adds one photo to a product from a public image URL. Shoots start from a product photo, so use it when a product has none suitable. Not available on Shopify stores: add photos to the product in Shopify instead.
Inputs: url (required), product_id (required).
Approves a ready frame: publishes the frame to the live product page on Shopify stores (on other stores it is saved to the product in Shootika). Shoppers can see it once it is published. Ask the person before approving, and show them the frame first.
Inputs: frame_id (required).
Returns the connected store (name and platform), its Shootika plan, and how many frames it has left this month (the quota). Use it first to learn what the store can do, or before starting shoots to check frames remain. Takes no arguments.
Inputs: no named parameters.
Returns one product with its id, title, family, garment_category, accessory_kind, shoot_routes (on_model: shoot it on a model scene; placement: on a placement scene, the product alone; empty: it cannot be shot yet) and photos (each photo has an id, url and size). Use it to pick which photo to shoot and which kind of scene to list, or to check a product before a shoot. Fails with "not found" for an id that is not in this store.
Inputs: product_id (required).
Returns one shoot (frame), waiting up to 8 seconds for it to finish (one wait at a time per connection; a second call answers at once). A ready frame comes with a small preview image and its full-size url. If it is still queued or processing when the wait ends, call get_shoot again with the same frame_id. A failed frame says why; its frame was refunded.
Inputs: frame_id (required).
Lists the AI models (the people shown wearing or carrying the product) this store can use: id, name, body type, skin tone, the poses each can take, its subject and its frames_per_shoot (the frames one shoot with it uses). A model with subject 'mannequin' is a ghost mannequin: the product shown hollow, with no person in it, on the studio styles only; a shoot with it uses 2 frames instead of 1. Use it with list_styles to choose a scene. Returns every model in one result; takes no arguments.
Inputs: no named parameters.
Lists the store's products, newest first, 24 per page, each with its id, title, family, garment_category, accessory_kind, shoot_routes (on_model, placement, both or none: the shoots it can have) and photos (the source images shoots start from). Use it to find a product id. Pass external_id to look one up by the store's own id. The result carries page, page_size and total; when more pages exist, call again with page + 1.
Inputs: page, external_id.
Lists the store's scenes, 24 per page, each with its id, kind, model, style, pose, status and a preview image. An on_model scene is a model in a style and pose; a lifestyle scene is a placement scene, the product alone in a style (model null, pose Placement). Use it to find a ready scene id for a shoot: pass kind lifestyle for a product whose shoot_routes include placement. Filter by kind, model_id, style, pose or status. The result carries page, page_size and total; when more pages exist, call again with page + 1.
Inputs: kind, page, pose, style, +2 more.
Lists the store's shoots (each result image is a frame), newest first, 24 per page: id, product, status, whether it is approved, and the result image url. The first 4 ready frames come with a small preview image and their full-size url as text; frames after the 4th carry their url in structuredContent only. Use it to check on a shoot or to find frames to approve. Filter by product_id, status or approved. The result carries page, page_size and total; when more pages exist, call again with page + 1.
Inputs: page, status, approved, product_id.
Lists the scene styles (backdrops and looks, plus the store's own reference scenes as ref:) a shoot can use: key, name, description and placement (true when the style also has a placement scene, for products shot alone). Use it with list_models to choose a scene. Returns every style in one result; takes no arguments.
Inputs: no named parameters.
Takes an approved frame back off its product: on Shopify stores it is removed from the live product page. The frame itself is kept, so it can be approved again until it expires. Ask the person before removing.
Inputs: frame_id (required).
Asks for a scene (a model in a style and pose) to shoot with. A library model on a preset style is free. A custom model's scene, or a scene on the store's own setting (a ref: style), uses 1 frame from the store's plan (an extra frame if the plan bills them and this connection may), refunded if it fails, so ask the person before requesting one they did not ask for. Safe to repeat: a scene the store already has comes back as it is, with no new charge. A new scene takes a little while to prepare; check list_scenes and start the shoot once its status is ready.
Inputs: pose (required), style (required), model_id (required).
Starts a shoot: one new frame of the product in the scene's style, on the scene's model, or the product alone for a placement scene (kind lifestyle). Each call uses one frame (two on a ghost mannequin) from the store's plan, so ask the person before starting shoots they did not ask for, and pass a request_id so a retry does not shoot twice. Pass notes to say how the product is worn or presented; notes never change the model or the scene. Returns at once with the frame queued; call get_shoot with its id to wait for the result.
Inputs: scene_id (required), product_id (required), source_image_id (required), notes, +3 more.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Virtual try-on and on-model AI fashion photography: catalog search, try-on grids, HD delivery.
Build and run AI product photography workflows: on-model shots, catalog images, prints and edits.
AI product photography for fashion sellers: Shopify photo audits, seasonal guides, AI try-on.
Create product photos, on-model images, ad visuals and short videos from one product photo.
Add one secure layer between your agents and this server.
