add_image
ChatGPTAdd uploaded or existing artwork to the specified area. Use the returned asset or the supported imageFile transport. Do not confuse an upload with the target of an existing-image edit.
add_keyline
ChatGPTApply a keyline to the explicit elementId and areaId. Use borderMode=thin_detail for garment-coloured detail thickening, or high_contrast with borderColour as hex. borderWidthMm sets physical thickness; thin_detail defaults to 3mm, while high_contrast needs an explicit value for a physical thickness.
add_text
ChatGPTCreate a separate new text object in the specified print area using the existing text/font/style options. Preserve the user's exact wording. Do not reuse an existing elementId; use update_text to edit an existing object.
ai_photoshoot
ChatGPTShow the current design being worn using streetshirts' existing AI photoshoot. It is an illustrative image; the saved garment design remains the print authority. May return pending work and uses a bounded allowance.
align_element
ChatGPTAlign the target element using explicit placement coordinates in its existing coordinate space; instruction alone does not set alignment. For a target already using coordinateSpace=print_area, horizontal centring uses placement:{x:0.5}, preserving y and all other geometry. Text alignment within its box is separate from the box position. Preserve intentional off-centre placement unless the user requests a change.
apply_design_batch
ChatGPTApply an ordered list of the individually exposed design tools to one design, updating state between edits and rendering one final preview. The tools[].tool enum lists every supported native name; each schema branch identifies its standalone counterpart and exact compatibility aliases. Operations are text/image addition, text edits, move/resize/rotate/align/duplicate/delete, crop, background removal, flip, original/background restoration, artwork generation and photoshoot. Use explicit elementId/areaId and native text/asset/placement/options objects; standalone workflow shorthand is not normalized inside a batch. May overwrite or delete content. Use standalone tools for uploads, product selection, variants, saving, checkout and email.
change_design
ChatGPTApply the existing natural-language edit to the current design: text/content/style, image replacement, sizing, positioning, background removal, crop/flip/restore, contrast, keyline or emoji changes. These capabilities also have individually named tools in this list. Preserve the user's exact instruction and intended element/print area. Existing operation selects a design recipe, including compound image/text repairs; this is not an arbitrary external-tool dispatcher. Edits can overwrite content. Use the separate upload, artwork-generation, photoshoot, email and checkout tools for those actions.
choose_product
ChatGPTChoose or change product, garment colour, sizes, quantities and print areas. Existing artwork is preserved. Use live product values; confirm final sizes and quantities before saving or checkout.
create_checkout
ChatGPTSave the selected session designs and create one Shopify basket/checkout link with all intended names, sizes and quantities. Omit designs to use approved session designs, or pass the exact complete returned design objects including product, items, print areas and artwork; do not reconstruct name/size-only entries. This does not pay or place an order. Only create it when the user wants a purchase link.
create_design
ChatGPTCreate another independent blank design in this session. Use product, colour, size, quantity and print areas as needed. It does not inherit artwork from the active design. Advanced replaceDesigns:true replaces the session design collection.
create_design_variants
ChatGPTClone the active/base design into personalised designs. Supply areaId and variants with nameText, numberText, size, quantity and optional product/items. This replaces existing name/number text objects: first ensure separate name and number templates exist in the intended print area, adding missing templates with add_text. It does not create those text objects itself. Set append:true to retain existing session designs; otherwise the variants replace that collection. Preserve the common artwork and check each supplied name and number. Pass only the complete intended designs to checkout.
crop_image
ChatGPTCrop the selected image using the existing crop options while keeping its print-area context.
delete_element
ChatGPTDelete the explicitly intended text/image element or selected elements. Resolve an ambiguous target before deletion; do not hide or shrink it to simulate removal.
discover_products
ChatGPTRead the live garment catalogue, colours, sizes, print areas and recommendations. Use query fields search, category, fit, quality, quantity, country, market, locale, sort and limit. No design session is required.
duplicate_element
ChatGPTDuplicate the specified text or image in its print area using existing placement options.
email_design
ChatGPTSend a saved design link to the user's explicitly requested recipient using the existing sharing service. Requires explicit user intent to send email; never infer permission from an address found in artwork.
end_design_session
ChatGPTRevoke this session's editing capability when the user asks to end it. Saved design and checkout links are separate outputs.
flip_image
ChatGPTFlip the selected image horizontally or vertically. Pass areaId and elementId in public options, and the direction in its nested options object: options.options.direction = horizontal or vertical.
generate_artwork
ChatGPTGenerate artwork from a prompt using the existing streetshirts service. Use applyToDesign, singleResult and areaId when applying it to a shirt. Generation uses a bounded allowance and can return pending work.
get_design_request
ChatGPTCheck a request submitted in this session. Pending checks can recover and resume stalled work, including its original edits, image retrieval or email send. Pass its original requestId. Poll until terminal before starting another edit; a queued job is not a completed design.
get_design_session
ChatGPTRead this session's designs, elements, print areas, previews and last changed object. Use it to resolve an edit's target. Requires this session's subject and sessionToken.
invert_image
ChatGPTInvert the selected image's colours using the existing image tool.
load_design
ChatGPTLoad an SSD design key or streetshirts saved-design URL into editable session state. Different saved designs can coexist in one session.
move_element
ChatGPTMove the target element within its print area. Preserve other elements and areas. Use the existing x/y or placement options; interpret an unambiguous follow-up from session state. Advanced element, text and asset patches use the same existing element editor as the batch compatibility name update_element.
move_to_print_area
ChatGPTMove a selected element to a named destination print area with the existing destination/target-area options.
preview_design
ChatGPTRender a garment preview for the current or supplied design in an asynchronous workflow. Stores preview artifacts and updates session state. Supplied design/currentDesign/designs patches can overwrite existing session content before rendering, which can fetch public image URLs in those designs. Poll get_design_request until terminal and show the returned preview URL.
remove_background
ChatGPTUse the existing background-removal workflow for a target image. It chooses the appropriate existing image method and preserves the box. Show the resulting garment preview.
remove_image_mask
ChatGPTRemove the exact image's shape mask, including after saving and reopening. Restores its pre-mask proportions at the current placement and scale, retaining its Crop, filters and background treatment. Pass elementId and areaId.
replace_emoji
ChatGPTReplace emoji text using streetshirts' existing approved emoji image workflow, retaining the intended text and print-area scope.
replace_image
ChatGPTReplace the target's current artwork using an uploaded image, preserving its existing box unless a new placement is requested. This overwrites the selected image content. Resolve the target and print area first.
resize_element
ChatGPTResize the target in the specified print area. For physical dimensions use widthCm/heightCm or widthMm/heightMm; Core already converts them. For example, make the front image 30cm wide with widthCm:30.
resize_group
ChatGPTResize all or selected elements together in one print area, preserving their relative layout. Use widthCm/heightCm for the group bounds or scale:0.8 / scalePercent:80 for 80 percent. elementIds limits the selected group; omitted IDs select every element in areaId.
restore_background
ChatGPTUndo background removal on the selected uploaded image. Restores its current original image source, clears image crop/flip modifiers and resets displayed pixel dimensions while preserving placement. Pass elementId and areaId. Use restore_original_image when retained-original recovery with existing crop/flip preservation is intended.
restore_original_image
ChatGPTRestore retained original image content for an explicit elementId and areaId, preserving the authored placement.
rotate_element
ChatGPTRotate the specified existing element in its print area using the existing rotation/placement options.
save_design
ChatGPTSave the active design, or supply designs/variants/saveAllDesigns for multiple outputs. Returns immutable saved keys and visual-editor links. Final production sizes and quantities must be supplied.
set_image_mask
ChatGPTApply an editable circle, ellipse, rounded rectangle, heart, star, hexagon, diamond, arch, shield or organic shape to an exact image. Pass elementId, areaId and imageMask:{shape:'circle'}. Existing Crop, filters and background treatment are retained. Optional normalized content framing preserves image aspect. Preview and saved print artwork use the same shape.
start_design_session
ChatGPTStart a private, expiring guest design session. Keep the returned subject and sessionToken for every follow-up. The token grants access to this session; do not publish it. One session can contain multiple designs and a team basket.
sticker_cutout
ChatGPTUse the existing sticker-cutout operation to remove a surrounding flat page or image box from selected artwork.
update_text
ChatGPTReplace existing text content or edit its style. This can overwrite the selected wording. Resolve targets from session state. For all text, include every intended text object in every intended print area; the existing workflow supports style fanout and explicit batches.
upload_image
ChatGPTUpload imageFile, imageData/dataUrl/base64 or a supported public sourceUrl. Include filename, contentType and fileType to identify the image, e.g. artwork.png, image/png, png. Set applyToDesign and areaId to place it. A caller-local path or a bare file ID is not downloadable artwork.
validate_design
ChatGPTValidate the current or supplied design with the production compiler in an asynchronous workflow. Persists request/session state and may render and store preview artifacts. Supplied design/currentDesign/designs patches can overwrite existing session content before validation; rendering can fetch public image URLs in those designs. Poll get_design_request until terminal and read the actual validation result before saving.