MCP App Store
by KeenEthics
Kainoteka icon

Kainoteka MCP Server for ChatGPT

By Kainoteka

Overview

Kainoteka is Lithuania's price comparison service. Ask for any product — "gaming headphones under 100€", "a work laptop with 64GB RAM up to 2000€" — and Kainoteka searches across 300+ Lithuanian online shops, shows product cards with photos and lowest prices, and compares offers shop by shop so you can see which shop sells it cheapest. Searches understand budgets, specs (RAM, storage) and product categories, and every result links to the full comparison on kainoteka.lt.

How to connect

  1. ChatGPT: Open the Kainoteka listing in the ChatGPT app directory and choose Connect. ChatGPT listing Kainoteka exposes 5 tools here.

Tools

get_categories

ChatGPT
Get the category tree of Kainoteka.lt (names in Lithuanian). Use the returned slugs as category_slug in search_products to narrow a search.

get_product_offers

ChatGPT
Get all shop offers for a single product found via search_products (use its slug). Returns every Lithuanian shop selling the product with prices in EUR, cheapest first, plus a short description and the product specs (RAM, CPU, storage, screen, …) — use the specs to verify requirements like "64GB RAM" before recommending. Use this to answer "where is it cheapest?". Only works for products with hasPage=true; for hasPage=false products there is a single offer already linked by their productUrl.

get_search_filters

ChatGPT
Get the brands, categories, shops and price range available for a search phrase or category. Use the returned slugs as brand_slug / category_slug in search_products.

resolve_category

ChatGPT
Call this FIRST when the user asks for a physical product type sold in online shops (laptop, phone, TV, headphones, …) — never for fuel, subscriptions, services, travel or other non-catalog price questions: it maps the product type to a Kainoteka category. Then call search_products with the returned category_slug and a query containing ONLY model names / specs (never the product type words). Pass the product type in Lithuanian when possible (e.g. "nešiojamas kompiuteris"). If no category is found, call search_products without category_slug using the best product-name query.

search_products

ChatGPT
Search for products across 300+ Lithuanian online shops via Kainoteka.lt price comparison. SCOPE: only physical consumer goods sold by online shops (electronics, appliances, home, sports, toys, tools, beauty …). Do NOT use for fuel/petrol prices, utilities, subscriptions or digital services (Netflix, Spotify), travel/flights/hotels, tickets, real estate, cars, groceries or restaurant prices — answer those without this app. Returns products with their lowest price in EUR and the number of shops selling them. Product names are mostly in Lithuanian — translate the user query into Lithuanian when possible (e.g. "gaming headphones" → "žaidimų ausinės"). Use max_price_eur / min_price_eur for budgets. WORKFLOW for product-type requests: call resolve_category first, pass its category_slug here, and put ONLY model names / specs in the query (e.g. query="64GB RAM", never "nešiojamas kompiuteris 64GB RAM"). If category_slug is omitted the server auto-detects the category and strips product-type words itself, reporting what it did — override with category_slug if wrong. Spec requirements like "64GB RAM" are verified server-side: names in storage context ("64GB SSD") are rejected, and top candidates are checked against their spec sheets (specVerified=true/false per product). Treat specVerified=false as unconfirmed. Each call searches ONE product or model — never list several models in one query (it matches nothing). The user sees product cards from your LAST search_products call only, so do not end with a narrow re-check search; make the final search the one whose results you present. To find the CHEAPEST offer for a specific model (e.g. "Samsung S24 256GB"), prefer sort=relevance and compare the returned prices (or use get_product_offers). BUDGET queries ("up to 2000€") set max_price_eur with sort=relevance and a larger per_page (12-20): the budget is a ceiling, not a wish for the cheapest — present options across the whole price range (budget / mid / high-end picks) with a recommendation. Use price_asc only when the user explicitly wants the cheapest. For broad "best X" requests use sort=popularity with per_page 12-20, recommend 2-3 options for different needs, and ask the user 1-2 clarifying questions (use case, budget) when the request is ambiguous. IMPORTANT: when linking to a product, copy its productUrl EXACTLY as returned. Never construct kainoteka.lt URLs from the slug — many products have no page on kainoteka.lt (hasPage=false) and their productUrl is a redirect to the shop; a constructed /product/{slug} URL returns 404.

Capabilities

Interactive

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/kainoteka"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/kainoteka)