App Store & Google Play Review Intelligence

What App Store & Google Play Review Intelligence does

App Store & Google Play Review Intelligence MCP server gives an MCP client read-only access to app listings and user reviews from the Apple App Store and Google Play. It can identify an app from a name, category, numeric Apple ID, Google Play package name, or supported store URL. Store selection is inferred from the identifier when possible.

The App Store & Google Play Review Intelligence MCP server is intended for questions about user sentiment and product changes. It can summarize the most common complaints, include representative review quotes, compare recent versions, and show whether negative feedback increased after an update. It also supports side-by-side comparisons of apps, including comparisons between iOS and Android listings.

How it works

The server exposes four MCP tools. search_apps finds candidate apps and returns identifiers, ratings, rating counts, prices, versions, genres, URLs, developers, and Google Play install counts. review_insights analyzes a sample of reviews and returns rating distributions, complaint phrases, sample quotes, and version deltas. Its latest-versus-previous-version analysis reports rating and negative-share changes plus complaints that are new or increasing.

get_app_reviews returns individual reviews with fields such as rating, title where available, body, version, date, helpful votes, and Google Play developer replies. compare_apps accepts two to ten app identifiers and reports store metrics, recent averages, positive and negative proportions, latest-version averages, leading complaints, and metric leaders.

Parameters include country or storefront codes, review pages, Google Play language, selected comparison versions, keyword limits, and quote settings. Review pages range from one to ten, with about 50 reviews per page. The tools are read-only, and failures are returned as MCP tool errors labeled with validation, upstream, or internal error prefixes.

Setup and configuration

For hosted use, connect an MCP client to the Streamable HTTP endpoint provided by the project and send an Apify API token as a Bearer authorization header. The README provides JSON configuration examples for Claude Desktop, Cursor, and Windsurf. Hosted calls use pay-per-event billing: searches, returned raw reviews, insight analyses, and comparisons have separate charges, with Google Play calls billed through corresponding event types.

For local use, install dependencies, build the project, and run the package through its stdio entry point. The documented commands are npm install, npm run build, and npx app-review-intel-mcp. The local option does not require App Store Connect or Play Console access according to the project documentation.

Tools and capabilities

• Search the Apple App Store or Google Play by app name or category.
• Analyze complaint frequency and supporting quotes.
• Compare feedback for the latest and previous app versions.
• Retrieve raw reviews for custom analysis, including Google Play developer replies.
• Benchmark two to ten apps, mixing Apple and Google Play identifiers.

As a 🔎 Search & Data Extraction listing, the App Store & Google Play Review Intelligence MCP server fits workflows that need review evidence rather than private analytics or store-console data.

Limitations and notes

All tools are read-only. The service does not provide App Store Connect or Play Console access. Results depend on reviews available from the relevant public store and the selected countries, pages, language, or version filters. review_insights and compare_apps return aggregates, while get_app_reviews is the tool to use when the complete available review text is needed.

Hosted usage requires an Apify token and incurs charges for successful calls or returned reviews according to the published event schedule. Validation failures and calls that cannot reach a store are not charged, as described in the README. Client support depends on MCP transport: the hosted deployment uses Streamable HTTP, while the local command runs over stdio.