Description

Himart for ChatGPT helps users search and explore products from Lotte Himart, a major electronics and home appliance retailer in Korea. Users can find items across categories such as TVs, refrigerators, laptops, and home appliances using natural language. Searches can be based on brand names, model numbers, product types, or specific preferences. The app provides product listing information available from Lotte Himart, such as product names, pricing details, and other relevant attributes when provided. The app helps users retrieve product information and explore available items from Himart’s catalog in a conversational way.

Website Preview

Screenshot of Himart website

App Screenshots

Himart screenshot

Capabilities

Works in Conversation

Publisher Intelligence

Insights and recommendations for app publishers. See how your app performs and how to improve discoverability.

Server Status himart-public-mcp-server v2.14.2

2
Tools
1
Resources
0
Prompts
https://mcp.ldccai.com/lottehimart

Last checked: 23h ago

Server Instructions

Lotte Himart Services MCP Server

Technical Details
Connection Latency 1.9s
30-Day Uptime 100.0%

Tools(2)

Showing 2 of 2 tools

Sorted by toolName
ToolDescriptionFlagsTestLast Tested
himart_product_options_retriever
Retrieves candidate OPTION codes (optCd) and COLOR codes (colorCd) to be used as filters for Himart product search. This is a **PRE-STEP tool** and DOES NOT return the final product list. NEVER use this tool as a standalone result. It MUST always be followed by `himart_products_retriever` to retrieve the final product list. [Disallowed service-based offerings – DO NOT CALL] - NEVER call this tool for service-only or entitlement-based offerings. - This includes (but is not limited to): gift cards, vouchers, coupons, subscriptions, rentals, care plans, extended warranties, installation services, delivery services, memberships, telecom plans, SIM/eSIM activation, or insurance products. - If the user request falls into these categories, DO NOT call this tool. [When to call] - You MUST call this tool BEFORE product search if the user request includes any attributes beyond simple price/brand/product-name conditions. Examples include: • numeric or capacity constraints (e.g., 800L+, 24kg, 65-inch) • type or form factor (e.g., 4-door, side-by-side, top-load) • features or performance (e.g., OLED, inverter, steam) • color/material/finish (e.g., white, black, stainless) • exclusion constraints (e.g., 'exclude white', 'no side-by-side') • exploratory queries (e.g., 'what options/colors are available?') [Mandatory usage flow] 1) Call this tool to retrieve option/color candidates. 2) Select the optCd and/or colorCd values that best match the user's constraints. - For range constraints (e.g., '800L or larger'), select ALL matching optCd buckets. 3) Immediately call `himart_products_retriever`, populating optCdFilter/optCdExFilter and colorCdFilter/colorCdExFilter with the selected codes. [Strict restrictions] - NEVER invent or infer optCd or colorCd values. - ONLY use codes returned by this tool. - Do NOT force options that are not present in the response. [Input rule] - query: A natural language description of the user's intent and constraints. It may include price, brand, specs, attributes, and colors, but MUST NOT add assumptions not stated by the user.
read-only
100%Latency 1.5s
May 9, 2026
himart_products_retriever
Returns the FINAL product list from Himart using hybrid retrieval (BM25 keyword search + vector semantic search). This tool NEVER generates or infers option codes (optCd) or color codes (colorCd). If valid codes are not available, set optCd/colorCd filters to None. Arguments must be wrapped under search_query object: { "search_query": { ...fields... } } [Disallowed service-based offerings – DO NOT CALL] - NEVER call this tool for service-only or entitlement-based offerings. - This includes (but is not limited to): gift cards, vouchers, coupons, subscriptions, rentals, care plans, extended warranties, installation services, delivery services, memberships, telecom plans, SIM/eSIM activation, or insurance products. - If the user request falls into these categories, DO NOT call this tool. [Direct use] - Call this tool directly ONLY when the user request is limited to: price/budget, brand include/exclude, or product name/model/category. [Option prefetch required] - If the user mentions any attributes that require option filtering (color, capacity/size, type, features, inch, OLED, ranges, etc.), DO NOT call this tool first. - ALWAYS call `himart_product_options_retriever` first, then call this tool with: {"search_query": { optCdFilter, optCdExFilter, colorCdFilter, colorCdExFilter }} (use only the filters that are needed). - Filter values MUST come ONLY from `himart_product_options_retriever` results. NEVER generate, guess, or infer any optCd/colorCd. - You MAY infer high-level attribute intent (e.g., '대화면', '신혼용', '거실용') ONLY to select the most relevant codes from the retriever result. - If no valid codes are available from the retriever, DO NOT call this tool. [Tool call construction rules] - ALWAYS construct arguments under: { "search_query": { ...fields... } } - REQUIRED on every call: userQuery, lexicalQuery, page, size, sort. - OPTIONAL filters (set only when applicable): minPriceFilter, maxPriceFilter, brndNmFilter, brndNmExFilter, optCdFilter, optCdExFilter, colorCdFilter, colorCdExFilter. - NEVER omit required fields. - NEVER place filters outside search_query. - NEVER include attributes or price expressions inside lexicalQuery. [Input rules] - **userQuery**: single-sentence natural language summary of the user's intent only (NO inference). In follow-up queries, preserve the previous core intent and constraints. - **lexicalQuery**: IDENTIFIER keywords only (max 4), extracted VERBATIM from the user's text (no added words). If the user provides a specific identifier (model/barcode/exact product code), lexicalQuery MUST contain ONLY that identifier. Otherwise, use ProductType/Category + (Brand OR Series/Line name). Do NOT append ProductType when a specific identifier is present. NOT allowed: any comparatives (e.g., above, below), attributes, specs, features, ranges, or price expressions. [Follow-up and Retry rules] - For follow-up queries, ALWAYS reuse the previous search_query as the base. Apply ONLY the newly mentioned changes (delta). Keep all existing queries, filters, sort, page, and size unchanged unless explicitly modified. - On retry due to schema/validation/parsing errors: Preserve ALL existing fields and filter values. Fix ONLY invalid field names, nesting, or types. NEVER drop existing filters, codes, or constraints unless forced by schema. - When re-calling after failure: Keep all previously valid optCd/colorCd filters. Do NOT remove filters unless the user explicitly requests relaxation. - You MAY re-call `himart_product_options_retriever` ONLY when: 0 results were returned AND attribute constraints were involved. [Price rules] - If user mentions price/budget/만원대 → set minPriceFilter/maxPriceFilter. - 'N만원대' = [N*10,000 , (N+1)*10,000*100 - 1] - '초반/중반/후반' = lower/middle/upper third of the band. - '약/내외' = ±10%. - If no price mentioned → None. [Range handling] - Range constraints may require MULTIPLE optCd values. Select ALL matching codes from the retriever result and apply them as OR filters. [Sort rules]- Use ONLY Himart sort codes: RK, LP, HP, SA, RV, NW. - NEVER use generic e-commerce sort labels such as BEST, PRICE_ASC, PRICE_DESC, NEW, REVIEW, SALES. - Map user intent to sort code:인기 / 추천 / 베스트→RK, 낮은가격→LP, 높은가격→HP, 판매량→SA, 리뷰 / 평점→RV, 신상품 / 최신→NW. - If the user does not specify sorting, keep sort as RK ( or keep previous sort on follow-up unless explicitly changed).
read-only
100%Latency 587ms
May 9, 2026

Discoverability Score

52

Fair

52 of 100 — how easily AI agents find your app

  • Description quality
    20/20
  • Example prompts
    0/20
  • Keyword coverage
    0/15
  • Tool metadata
    16/20
  • Visual assets
    5/20
  • Endpoint health
    10/10
  • Data freshness
    11/15

How to Improve

Add at least 2 example prompts. Prompt examples strongly improve app matching and click-through intent.

Increase keyword coverage (discovery + trigger) to improve retrieval for long-tail queries.

Provide a stable HTTPS logo URL (avoid connectors://) so cards render consistently across clients.

Add at least 2 screenshots that show real workflows to increase confidence and conversion.

Read the full discoverability guide →

Technical Details

Status
ENABLED
Type
AI-Powered App
Auth
Open Access
Listed on
ChatGPT
Added
March 4, 2026
Last synced
May 3, 2026
Last checked
23h ago
Version
2.14.2
Distribution
Ecosystem Directory

Related Apps in shopping