What does the Daiso Cli AI skill do?

Use this when a user wants to search Daiso/다이소, compare product price candidates, nearby restaurants/cafes/음식점/카페, convenience stores/편의점, marts, Olive Young/올리브영, Megabox/메가박스, Lotte Cinema/롯데시네마, or CGV data through the Daiso project. Honor explicit no-install requests with connected MCP tools first, then HTML GET via an available browser, then web reading even when a shell is available; otherwise prefer the daiso CLI for direct execution, use the MCP endpoint when the host app supports remote MCP, and choose commands for products, stores, inventory/재고, compare, places, movies, showtimes,...

Why use Daiso Cli on TypingMind?

Because you install it once and use it with any model. Daiso Cli is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Daiso Cli in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/hmmhmmhm/daiso-mcp/tree/main/skills/daiso-cli. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Daiso Cli?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Daiso Cli?

As many as you like. As long as a model supports skills, you can use Daiso Cli with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Daiso Cli AI skill free?

Yes. It is published on GitHub by hmmhmmhm under the MIT license. You only pay your own AI provider for the tokens you use.

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

Daiso Cli by hmmhmmhm — AI Skill for Any LLM

Daiso CLI

Use this skill to operate the Daiso MCP project through npx daiso and the public MCP endpoint.

설치 없이 첫 조회

최신 서버 안내는 현재 버전과 최신 연결에서 제공합니다. 웹 도구는 과거 리디렉션·수집본을 반환할 수 있으므로 표시된 버전과 HTML 생성 시각을 확인하세요.

연결된 MCP 도구가 있으면 search_products에 {"service":"daiso","query":"수납박스","pageSize":3}를 전달합니다.
실제 브라우저 도구가 있으면 /browse/ HTML GET 주소를 열고 실제 상품·매장·재고를 읽습니다. Work의 내장 브라우저도 사용 가능합니다.
웹 읽기(open/WebFetch)만 가능하면 웹 조회 시작의 실제 링크 ID로 click합니다. 글자 입력·상품·매장 지역 선택 링크를 따르고 새 URL을 조립하지 마세요. 이전 HTML이면 같은 조건으로 새 응답 확인 링크를 최대2번 선택합니다. TimeoutError는 같은 실제 링크를 한 번만 재확인하고 계속 실패하면 미확인으로 보고합니다. 새 HTML도 기존 API 캐시를 사용하며 원본 갱신을 보장하지 않습니다. click이 없는 도구의 동일 동작은 보장하지 않습니다.
위 세 수단이 없거나 모두 실패하면 실제 조회 제약과 오류를 알립니다. 셸이 있어도 설치 없는 웹 조회가 우선이며 터미널·셸로 우회하지 않습니다. 셸 DNS 오류를 웹 GET 실패로 확대하거나 실행하지 않은 요청의 실패를 만들지 마세요.

https://mcp.aka.page/browse/read

웹 도구의 Crawled: today만으로 현재 재고라고 단정하지 마세요. 응답의 HTML 생성 시각과 실제 수량을 확인하고, 수량이 null이면 품절이 아닌 미확인으로 보고합니다.

응답의 글자 선택 링크로 사용자 검색어를 입력하고 상세·매장·재고도 실제 링크 ID로 선택하세요. 모든 후속 웹 GET은 /browse/ HTML입니다. 자동 HTML이며 쿼리 문자열은 필요 없습니다. 첫 조회 전에 npx·패키지·스킬 설치, health/help, 전체 도구 목록이나 OpenAPI 탐색을 요구하지 않습니다. REST GET 성공과 MCP 도구 호출 성공은 구분합니다.

상품 이미지는 해당 조회 응답의 imageUrl만 사용합니다. 전체 URL과 query string을 유지하고, 없거나 표시되지 않으면 생략하며 image_query·이미지 검색·검색 캐러셀을 사용하거나 웹 검색 이미지·다른 상품·다른 국가 이미지로 대체하지 않습니다. 실제 도구·URL·반환 데이터·오류만 보고하고 실행하지 않은 요청의 성공·실패를 만들지 않습니다.

공개 MCP는 https://mcp.aka.page, Streamable HTTP, 인증 없음입니다. 연결은 표준 initialize 절차를 사용합니다. 공개 도구 10개는 늘리지 않습니다. 기존 서비스별 이름은 deprecated 별칭이며 새 호출에 쓰지 않습니다.

Core Rule

무설치 요청이 아니고 CLI 사용이 허용된 환경에서는 CLI를 우선 사용한다. The CLI is the preferred path when you can run shell commands and package execution is allowed. Use the MCP endpoint when the user is configuring an AI app or explicitly asks for MCP connection details.

Quick Checks (오류 진단이 필요할 때만)

bash
npx daiso health
npx daiso url
npx daiso help

MCP server URL:

text
https://mcp.aka.page

Use --json when the user needs structured data, when comparing results, or when another tool will consume the output.

Common Commands

bash
npx daiso products 수납박스 --json
npx daiso stores 강남역 --limit 5 --json
npx daiso inventory 1034604 --keyword 강남역 --json
npx daiso display-location 1034604 04515 --json
npx daiso compare 콜라 --limit 3 --json
npx daiso places 강남역 --category cafe --limit 5 --json
npx daiso places 성수동 --keyword 브런치 --limit 5 --json
npx daiso get /api/feedback/requests --type bug --title "올리브영 재고 오류" --description "check_inventory(service=oliveyoung)가 빈 결과를 반환합니다." --service oliveyoung --toolName check_inventory --json
npx daiso gs25-products 콜라 --limit 10 --json
npx daiso gs25-stores 강남 --limit 10 --json
npx daiso gs25-inventory 오감자 --storeKeyword 강남 --storeLimit 10 --json
npx daiso seveneleven-products 삼각김밥 --size 10 --json
npx daiso seveneleven-stores "안산 중앙역" --limit 10 --json
npx daiso seveneleven-inventory 핫식스 --storeKeyword "안산 중앙역" --storeLimit 10 --json
npx daiso emart24-products 커피 --pageSize 10 --json
npx daiso cgv-movies --playDate <YYYYMMDD> --theaterCode <theaterCode> --json
npx daiso cgv-timetable --playDate <YYYYMMDD> --theaterCode <theaterCode> --json

For more command selection examples, read references/cli-command-map.md.

Request Recipes

아래 CLI 레시피는 CLI 사용이 허용된 경우에만 적용합니다. 무설치 요청에서는 위 순서대로 연결된 MCP 또는 웹 읽기의 HTML GET으로 실행합니다.

"콜라 어디가 싸?", "컵라면 가격 비교해줘": run npx daiso compare <keyword> --json first. Confirm stock or sale prices with service-specific inventory only when the user asks for store-level availability.
"강남역 근처 카페", "성수동 브런치 음식점": run npx daiso places <location> --category cafe|restaurant --json or use --keyword for the food or mood. Do not ask which retail brand they mean.
"다이소 핫식스 재고", "올리브영 두바이초콜릿 있어?": keep the named brand. Search that service first even if the product seems unusual for the brand, then suggest alternatives only after no results.
"GS25 강남 오감자 재고": separate product and location. Search product candidates first when an item code is needed, then run inventory with --storeKeyword.
"오늘 강남 CGV 시간표": compute today in KST as YYYYMMDD, find the theater if the code is unknown, then query timetable.
"이 오류나 개선 요청을 개발자에게 보내줘"처럼 명시적으로 요청한 경우에만 use submit_request with requestType=developer and a request object when MCP is available. In CLI-only environments, use npx daiso get /api/feedback/requests --type bug|improvement|feature|docs --title "..." --description "..." --json.
If a command fails, retry once with narrower input. If it still fails, report the service, command, and condition instead of inventing results.
If output contains imageUrl, preserve the full URL including query string when showing images.

Multi-step Korean request patterns

Convenience store product near a place: if the request has both product and location, prefer inventory lookup over product-only search. Example: npx daiso gs25-inventory 콜라 --storeKeyword 강남역 --storeLimit 10 --json.
Cross-service price candidate comparison: if the user asks which service is cheaper without naming a brand, use npx daiso compare <keyword> --json first. This uses no new external API key and compares Daiso, GS25, Seven-Eleven, and Emart24 product search candidates.
Nearby restaurants or cafes: use npx daiso places <location> --category cafe|restaurant --json. If the user gives a specific food or mood, use --keyword, for example npx daiso places 성수동 --keyword 브런치 --json.
Seven-Eleven inventory: use npx daiso seveneleven-inventory 핫식스 --storeKeyword "안산 중앙역" --storeLimit 10 --json.
Daiso inventory by product name: search products first, keep the selected product ID, then run inventory with a store keyword. 위치가 없으면 ask the user for an area or store before checking inventory.
Cinema movies and timetable: find the theater first when the theater code is unknown, then call movies or timetable. If the user says today or omits a date, compute today in KST as YYYYMMDD; do not copy example dates.

Workflow

Lotte Mart is discontinued; MINISO is unsupported. Japan DAISO supports products/stores/online availability only, not store inventory.

Identify the target service from the user request: feedback/developer request, compare, Daiso, places/restaurants/cafes, GS25, Seven-Eleven, CU, Emart24, Olive Young, Megabox, Lotte Cinema, or CGV.
Choose a CLI command from the common commands or references/cli-command-map.md.
Add --json for machine-readable output or when summarizing multiple records.
If a CLI command is not available for the exact route, use npx daiso get /api/... --json.
If a command fails, run npx daiso health and retry with a narrower query or service-specific endpoint.

Fallbacks

npx 또는 Node.js를 사용할 수 없으면 use connected MCP tools or start web reading at /browse/read and click the actual response link IDs. Do not construct new URLs. Use query-free /browse/ paths for every web GET. HTML is automatic. If neither is available, report the actual limitation without inventing results.
For network failures, retry once with a narrower query, then report the failing service and command.
For no results, preserve the command used and suggest a broader keyword or nearby area.
Switch from CLI to MCP only when the user is configuring an AI app, the shell cannot run npx, or the host environment already exposes the MCP server.

Output Handling

Summarize only the fields the user needs: product name, price, store name, address, stock, movie title, showtime, seat count.
Preserve IDs such as product IDs, store codes, theater codes, and movie codes when the user may need follow-up lookup.
Mention when data comes from live external services and may change.
In shell commands, quoted Korean strings are fine for values with spaces. URL-encode only when writing raw URLs.

MCP Usage

For AI apps that support remote MCP, tell the user to connect:

text
https://mcp.aka.page

For local shell work, prefer npx daiso because it gives direct CLI commands and JSON output without requiring MCP host configuration.

Publisher	hmmhmmhm
Repository	`daiso-mcp`
Skill name	`daiso-cli`
Stars	346
Forks	29
Bundled files	1
License	MIT
Links	Skill source Repository Homepage

Daiso Cli

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Daiso CLI

설치 없이 첫 조회

Core Rule

Quick Checks (오류 진단이 필요할 때만)

Common Commands

Request Recipes

Multi-step Korean request patterns

Workflow

Fallbacks

Output Handling

MCP Usage

Bundled files

Frequently asked questions

Set up your own AI workspace now

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Daiso CLI

설치 없이 첫 조회

Core Rule

Quick Checks (오류 진단이 필요할 때만)

Common Commands

Request Recipes

Multi-step Korean request patterns

Workflow

Fallbacks

Output Handling

MCP Usage

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now