@perttu/app-store-mcp

@perttu/app-store-mcp

MCP server for looking up apps, reviews, ratings, charts, in-app purchases, privacy details, and version history on Apple's App Store. It requires no API key.

Category
访问服务器

README

@perttu/app-store-mcp

An MCP server for looking up apps, reviews, ratings, charts, in-app purchases, privacy details, and version history on Apple's App Store. It is built on @perttu/app-store-scraper and does not require an API key.

Requirements

  • Node.js 20 or newer
  • An MCP client that can launch local stdio servers

Setup

You do not need to install the package globally. Add this to your MCP client's server configuration:

{
  "mcpServers": {
    "app-store": {
      "command": "npx",
      "args": ["-y", "@perttu/app-store-mcp"]
    }
  }
}

Restart the client after changing its configuration. It should discover eleven tools whose names begin with app_store_.

If your client cannot find npx, replace npx with its absolute path. Run which npx on macOS or Linux, or where npx on Windows, to find it.

To install the executable yourself instead:

npm install --global @perttu/app-store-mcp

Then use app-store-mcp as the command and omit the arguments array.

Tools

Tool What it returns
app_store_app Full metadata for one app, addressed by track ID or bundle ID
app_store_search Keyword search results, with pagination and an IDs-only option
app_store_list Ranked App Store collections, optionally filtered by category
app_store_developer Apps published by a developer
app_store_reviews Recent or helpful reviews, up to page 10
app_store_ratings Total rating count and the 1–5 star distribution
app_store_similar Apps related to a given app
app_store_suggest Autocomplete suggestions for a search term
app_store_privacy Privacy labels and policy information
app_store_version_history Previous versions, dates, and release notes
app_store_in_app_purchases Top in-app purchase names and localized prices

Most tools accept a two-letter country code. The default is us. Tools backed by Apple's lookup and search APIs also accept a lang value such as en-us.

Every tool is read-only. Results are returned as both structured content and formatted JSON text, so they work with clients that support either response form.

Arguments

country is optional and defaults to us. Use a two-letter store code such as fi, gb, or jp; uppercase codes are accepted too. lang is an optional locale such as en-us.

Tool Required Optional
app_store_app id or appId country, lang, ratings
app_store_search term country, lang, num (1–200), page (from 1), idsOnly
app_store_list — country, lang, collection, category, num (1–200), fullDetail
app_store_developer devId country, lang
app_store_reviews id or appId country, lang, page (1–10), sort
app_store_ratings id country
app_store_similar id or appId country, lang
app_store_suggest term country
app_store_privacy id country
app_store_version_history id country
app_store_in_app_purchases id country

id, devId, and category are numeric. appId is the bundle identifier, for example com.apple.Pages.

The available collections are:

topmacapps                 topfreemacapps           topgrossingmacapps
toppaidmacapps             newapplications          newfreeapplications
newpaidapplications        topfreeapplications      topfreeipadapplications
topgrossingapplications    topgrossingipadapplications
toppaidapplications        toppaidipadapplications

Review sort is either mostRecent (the default) or mostHelpful. category is one of Apple's numeric App Store genre IDs; common values include games (6014), education (6017), productivity (6007), utilities (6002), and business (6000). The complete allowed set is included in the tool's input schema and exported as category from the package.

Example requests

The exact prompt is up to the client. These are representative:

Find the first five App Store results for "interval timer" in Finland.

Show the rating breakdown and recent reviews for app 553834731.

What data does app 284882215 say it collects?

app_store_list can fetch full metadata for every chart entry with fullDetail: true. That makes one additional App Store request per result, so use it with a small num value.

Checking the server

The MCP Inspector can launch the published package and call its tools directly:

npx @modelcontextprotocol/inspector npx -y @perttu/app-store-mcp

Running from source

git clone https://github.com/plahteenlahti/app-store-mcp.git
cd app-store-mcp
npm ci
npm run build
node dist/cli.js

The last command waits for MCP messages on stdin; it does not open an interactive prompt. For a local client configuration, point command at node and pass the absolute path to dist/cli.js in args.

Useful development commands:

npm test
npm run check
npm run build

Agent skill

The repository includes a concise Agent Skills workflow at skills/app-store-research. It teaches agents to choose the narrowest tool, reuse identifiers, and avoid unnecessarily large responses.

Codex and Claude Code discover it automatically while working in this repository through .agents/skills and .claude/skills. To use it in other projects, copy the canonical skill folder to your user skill directory:

# Codex
cp -R skills/app-store-research ~/.agents/skills/

# Claude Code
cp -R skills/app-store-research ~/.claude/skills/

Notes

This project reads public Apple endpoints. Apple can change those endpoints or apply rate limits without notice. The server deliberately does not cache or throttle requests; the calling client remains in control of when tools run.

This project is not affiliated with or endorsed by Apple Inc. App Store is a trademark of Apple Inc.

License

MIT

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选