Appstro MCP Server

Give your AI agent App Store superpowers. The Appstro MCP server exposes 8 free tools — app search, top charts, download & revenue estimates, publisher portfolios, reviews and ad intelligence — to Claude, Cursor, or any client that speaks the Model Context Protocol. No API key, no signup.

Connect in one command

Claude Code
claude mcp add --transport http appstro https://www.appstro.dev/api/mcp
Cursor / other MCP clients (mcp.json)
{
  "mcpServers": {
    "appstro": {
      "url": "https://www.appstro.dev/api/mcp"
    }
  }
}

On claude.ai: Settings → Connectors → Add custom connector → paste https://www.appstro.dev/api/mcp

8 tools included

search_apps

Search the App Store by keyword — name, developer, category, price, rating, rating count.

get_app

Full app details by trackId, including monthly download & revenue estimates.

top_charts

Top free / grossing / paid charts for any country.

app_estimates

Batch download & revenue estimates for up to 100 apps at once.

search_publishers

Find publishers by name with app counts and total ratings.

publisher_apps

A publisher's full portfolio with per-app revenue estimates.

app_reviews

Recent customer reviews for any app in any country.

app_ads

Google Ads Transparency: which ads an app publisher is running right now.

Ask your agent things like

  • Which meditation apps make over $100K/month with under 5,000 ratings?
  • Compare Duolingo and Babbel's portfolios: downloads, revenue and ratings.
  • What are users complaining about in the latest reviews of my competitor?
  • Is Calm running Google ads right now, and for how long?
  • Show me the top grossing apps in Japan and estimate their revenue.

How it works

The server implements the MCP Streamable HTTP transport at https://www.appstro.dev/api/mcp and is fully stateless — no sessions, no auth. Under the hood it queries the same public data sources that power appstro.dev: Apple's iTunes APIs, public download/revenue estimates, and the Google Ads Transparency Center. Please use it fairly; heavy automated scraping may be rate-limited.