Aso

npm version License: MIT Node.js MCP

App Store Optimization toolkit for AI assistants. Keyword research, competitor analysis, review sentiment, and metadata optimization. All through the Model Context Protocol.

No API key required. Works out of the box with real App Store data. Supports 155+ countries.

Quick Start

npx aso-mcp

Or install globally:

npm install -g aso-mcp

Why aso-mcp?

• 20 specialized ASO tools: from keyword discovery and ranking trends to App Store Connect metadata management
• Real App Store data: live search results, ratings, reviews, and suggestions
• Custom scoring engine: proprietary algorithm independent of Apple Search Ads API issues
• No API key needed: zero configuration, install and go
• Smart caching: SQLite-backed cache for fast repeated queries
• Rate limiting: built-in request management to avoid Apple throttling
• Multi-country: analyze keywords across 155+ App Store markets

Integration

Claude Desktop

Add to your config file:

{
  "mcpServers": {
    "aso-mcp": {
      "command": "aso-mcp"
    }
  }
}

{
  "mcpServers": {
    "aso-mcp": {
      "command": "npx",
      "args": ["tsx", "/ABSOLUTE/PATH/TO/aso-mcp/src/server.ts"],
      "cwd": "/ABSOLUTE/PATH/TO/aso-mcp"
    }
  }
}

Claude Code

claude mcp add -s user aso-mcp -- npx aso-mcp

Other MCP Clients

Any MCP-compatible client (ChatGPT, Cursor, Windsurf, etc.) can connect via stdio transport. Point it to the aso-mcp command.

Tools

Phase 1: Keyword Research

Phase 2: Competitor Analysis & Optimization

Phase 3: Localization & Reporting

Phase 4: ASO Generation

Phase 5: App Store Connect

Directly read and update your app's metadata on App Store Connect without leaving the AI assistant.

export ASC_ISSUER_ID="your-issuer-id"
export ASC_KEY_ID="your-key-id"
export ASC_PRIVATE_KEY_PATH="/path/to/AuthKey_XXXXX.p8"

"Set up App Store Connect with issuer ID xxx, key ID yyy, and key at /path/to/AuthKey.p8"

Utility

Usage Examples

Just ask your AI assistant naturally:

"How competitive is the 'fitness' keyword in the US?"

"Analyze Spotify's competitors and find keyword opportunities"

"Generate an ASO report for com.spotify.client"

"Compare 'music' and 'podcast' keywords across US, UK, and DE markets"

"Do a keyword gap analysis: Spotify vs Apple Music"

"Analyze Shazam's user reviews"

"Suggest title and subtitle for my fitness app targeting: workout, training, exercise"

"Discover keywords for a new calorie tracking app"

"Update my app's subtitle to 'AI Workout Planner' on App Store Connect"

"Show all locales and metadata status for my app"

Scoring Algorithm

The server calculates its own scores, independent of Apple Search Ads API:

When the aso npm package fails to reach Apple (503 errors), the server automatically falls back to custom scoring using search result analysis. Scores are always available.

Development

git clone https://github.com/kenanatmaca/aso-mcp.git
cd aso-mcp
npm install

npm run dev          # Run with tsx (development)
npm run build        # Compile TypeScript
npm run inspect      # MCP Inspector UI

# Tests
npx tsx test.ts              # Core tests (17)
npx tsx test-phase3.ts       # Localization & report tests (4)
npx tsx test-generation.ts   # ASO generation tests (8)

Tech Stack

• TypeScript + Node.js 22+
• MCP SDK: Model Context Protocol
• app-store-scraper: App Store data
• aso: ASO scoring with automatic fallback
• better-sqlite3: Cache layer
• Zod: Schema validation

License

MIT

Author

Kenan Atmaca