hmmhmmhm/daiso-mcp

다이소·편의점 재고와 생활 정보를 AI에서 바로 조회하세요. 설치·API 키 없는 공개 MCP, 웹 링크 조회, CLI·Skill을 지원합니다.

Daiso MCP – AI‑ready retail & cinema data service

What it is – A publicly hosted Model‑Context‑Protocol (MCP) server that exposes Korean retail, location and cinema information (Daiso, GS25, 7‑Eleven, CU, Olive Young, CGV, Megabox, Lotte Cinema, etc.) as structured API calls. The service can be accessed without any API key or local installation – AI agents, chat‑bots or developers simply point their MCP client at https://mcp.aka.page and invoke the provided tools.

Why it matters for AI – The data is packaged in the MCP standard, which many LLM‑based agents (Claude, ChatGPT, Home Assistant, Grok, etc.) understand natively. This lets a conversational model ask questions like “다이소 mcp로 수납박스 검색해줘” and receive a ready‑to‑use JSON response containing product names, prices, store locations and stock levels. No custom scraping or credential handling is required.

Core capabilities

Category Services (source) Main queries
Place Naver local search Nearby restaurants, cafés, desserts, other points of interest
Transport 오피넷 (Oil‑price network) National average fuel price, cheapest stations, station‑by‑station details
Retail comparison Daiso, GS25, 7‑Eleven, 이마트24 Price‑comparison for the same product across stores
Store data Daiso, Olive Young Product catalog, store list, stock availability
Convenience‑store data GS25, 7‑Eleven, CU, 이마트24 Product list, store list, stock, popular search terms, catalog
Cinema CGV, Megabox, Lotte Cinema, D‑trix (indie theatres) Theaters, movies, showtimes, remaining seats

All endpoints follow a standard MCP response model (standard.products, standard.stores, standard.theaters).

How to use it

1. Direct MCP connection (no install)

  1. In your AI app, add a remote MCP server with URL https://mcp.aka.page and transport HTTP / Streamable HTTP.
  2. Ask the model a natural‑language request, e.g.:
    다이소 mcp로 수납박스 검색해줘
  3. The agent calls the tool daiso_search_products with {"query":"수납박스"} and returns a structured list of matching items and prices.

2. HTTP POST example (any terminal or tool that can send JSON‑RPC)

curl -sS -N https://mcp.aka.page/ \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json, text/event-stream' \
  --data '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"daiso_search_products","arguments":{"query":"수납박스","pageSize":1}}}'

A successful response contains result.structuredContent (or result.content) with the product data.

3. CLI (optional) – npx daiso

If you prefer a local command line, run:

npx daiso products 수납박스 --json

The CLI talks to the same remote MCP server, but you can also self‑host the worker (see Contributing).

4. Codex Skill

For agents that can execute CLI commands, install the skill via ClawHub:

clawhub install daiso-cli

The skill maps natural‑language intents to the appropriate daiso CLI invocations.

Limits & reliability

  • Public GET API for Olive Young, CGV, CU, GS25 is capped at 3 000 calls per IP per day (KST). Exceed the limit → host your own instance (the repo is MIT‑licensed).
  • Detailed quota tables and health‑check docs are in docs/health-check-stock-and-quota.md and scripts/ops/README.md.
  • Real‑time service status: https://aka-page.betteruptime.com/.

Getting started locally (if you need your own server)

  1. Clone the repo.
  2. Deploy the Cloudflare Worker daiso-mcp (the project ships a wrangler.toml).
  3. The server runs on the same MCP protocol, so you can point your AI client at your own URL.

Documentation & further reading

  • Client guide – how apps, CLI, Skill and OpenAPI connect: docs/client-guide.md
  • Service reference – full list of tools, parameters and response schema: docs/service-reference.md
  • AI instruction – recommended prompting flow for agents: docs/ai-instruction.md
  • Web discovery – how the public front‑page works: docs/web-discovery.md
  • Contribution guide – building, testing and deploying: CONTRIBUTING.md

TL;DR

  • Daiso MCP = a ready‑to‑use, key‑less MCP server exposing Korean retail, fuel‑price and cinema data.
  • Connect your LLM‑agent to https://mcp.aka.page (or self‑host) and ask natural‑language queries; the model receives structured JSON results.
  • Optional CLI (npx daiso) and Codex Skill (clawhub install daiso-cli) give the same functionality in terminals or agent toolkits.
  • Free for casual use; higher traffic requires self‑hosting (MIT‑licensed).

License: MIT Hosted on: Cloudflare Workers (badge shown in README) Status: Live, with uptime monitoring via Better Stack.

Related

  • Project
    marc-shade/world-intel-mcp

    World‑Intelligence MCP Server is an open‑source Python platform that gathers free public‑API data from 30+ domains (finance, geopolitics, climate, military, cyber, health, etc.), stores it in SQLite and a Qdrant vector store, and exposes it via the Model Context Protocol for AI agents. It includes a Streamlit dashboard, CLI, PDF/HTML reporting, geofence alerts, and cross‑domain analytics, all built around 139 “tools” that fetch and analyse the data.

  • Project
    arabold/docs-mcp-server

    Grounded Docs MCP Server is an open‑source, locally‑run service that scrapes and indexes documentation (websites, GitHub, npm, PyPI, local files) for any library version, stores it (optionally with vector embeddings), and exposes a MCP‑compatible HTTP endpoint. AI coding assistants (Claude, Copilot, Gemini, etc.) can query the server to get exact, up‑to‑date answers, reducing hallucinations while keeping data private.

  • Project
    tavily-ai/tavily-mcp

    Tavily MCP is a Model Context Protocol server that exposes four web‑knowledge tools—search, extract, map, and crawl—so LLM‑based agents (Claude Code, Cursor, etc.) can fetch up‑to‑date web data, pull page content, generate site maps, or bulk‑crawl sites. It can be used via Tavily’s hosted endpoint or run locally with a single `npx` command, authenticating with an API key or optional OAuth. Configuration is done through environment variables (`TAVILY_API_KEY`, `DEFAULT_PARAMETERS`, `TAVILY_HUMAN_ID`). The server enables real‑time web access for AI assistants, RAG pipelines, and productivity tools without custom scraping code.

  • Project
    chrisryugj/korean-law-mcp

    A Model Context Protocol (MCP) server that connects AI assistants to official South Korean legal databases to provide accurate, verified legal research and legislative analysis.

  • Project
    tae0y/real-estate-mcp

    An MCP server that connects AI assistants to South Korea's MOLIT real estate API to provide live transaction data and financial investment simulations.