supershopping

Analyze price trends for a SuperShopping category, including price distribution, lowest-price stores, and alternatives. Use when the user asks about market prices or fair prices (e.g., 'how much are graphics cards these days', 'laptop price range'). Category can be a name (e.g., 'laptop') or ID. Returns price statistics (min/median/max and distribution), lowest-price stores, and alternative products.

Compare multiple SuperShopping products (2-5) with a spec and price comparison matrix plus a one-line conclusion. Use when the user wants to compare several candidates (e.g., 'which is better among these?'). Get product_ids from search results. Returns attributes (common keys), matrix (values/prices per product), and verdict (conclusion in Korean).

Get full details of a SuperShopping product, including specs, prices by store (offers), and new/used cross-comparison. Use when the user wants to examine a product in depth or is about to purchase. Obtain product_id from search/recommend results. Returns full attributes, offers with buy_url per store, comparison_url, other_sources (other sources of the same model, new vs used price spread), and affiliate_disclosure if present. Use other_sources to inform about price differences such as 'how much is the used version of the same model'.

Return the representative image of a SuperShopping product as an actual renderable image. Use when the user wants to see a product photo (e.g., 'show me this coat'). Thumbnails from search/recommend results are URLs (text) and may not display, so use this tool to fetch the image itself when visual confirmation is needed. Returns image content; do not show multiple at once, focus on products of interest.

Connect to purchase a selected product, returning the lowest-price store's purchase link, price, and disclosure. Use when the user expresses intent to buy (e.g., 'I want to buy this', 'purchase'). Actual payment occurs on the merchant's page. Returns payload with merchant, buy_url (affiliate-converted), price, etc., affiliate_disclosure if present, and next_actions. Present buy_url as a clear purchase link and always disclose affiliate_disclosure if present. For used items (personal transactions), guide the user on safe payment and physical verification.

Recommend a coordinated outfit based on a seed product or intent. Use when the user wants a full outfit composition (e.g., 'coordinate with this coat', 'fall date look'). Arguments: product_id (seed item, optional), intent (purpose/mood, optional), budget (overall budget cap, optional). At least one of product_id or intent is required. Returns seed, items (each with slot, title, min_price, reason, buy_url, thumbnail, source), styling_reason, and affiliate_disclosure if present. Present each item with a buy_url link and show photos using get_product_image.

Recommend products based on intent and budget, with reasoning. Use when the user provides only purpose/taste/budget without a specific search query (e.g., 'recommend a winter date look', 'budget laptop around 500k'). Arguments: intent (purpose/taste), budget (upper limit), must_have (required conditions). Returns recommendations with title, min_price, reason, buy_url, thumbnail, source, and affiliate_disclosure if present. Present each recommendation with a buy_url link and show photos using get_product_image.

Search new and used products using natural language (the main entry point of SuperShopping). Use when the user is looking for a specific product. Automatically extracts domain, gender, price range, and attributes from the query for filtering. Returns items with title, min_price, key_specs, thumbnail, buy_url, source, condition, etc. Always present each buy_url as a link. Use get_product_image to render photos. If the user only wants recommendations or the query is ambiguous, use recommend_products.