wtf is google-play-analytics-mcp?

This project is a connector, called an MCP server, that gives AI assistants like Claude access to Google Play Console analytics for an Android app. Google's Play Console shows useful data such as how many people found an app by browsing versus searching the store, install and uninstall counts, ratings, crash reports, subscriptions, and reviews, but there is no easy way for an AI agent to read that data directly. This project solves that by reading the CSV report files that Google Play automatically exports to a Cloud Storage bucket, and turning them into a set of tools an AI can call. It connects using a Google Cloud service account, so there is no manual browser login or token refresh needed. Because Google Play publishes these reports with a one to two day delay, the freshest data available is usually from a day or two ago. The tools it exposes cover things like acquisitions by traffic source or country, store listing visitor to install conversion rates, search terms that bring in visitors, install and uninstall counts, average ratings, crash and error reports, subscription activity, and individual user reviews. There is also a general purpose tool for pulling raw report data for anything not covered by the specific tools. Setting it up requires a Google Play Console account with reports already being exported, a Google Cloud service account with read access to that report bucket, and Python 3.11 or newer. Configuration is handled entirely through environment variables, and the service account key is never meant to be committed to the repository. This project is aimed at Android app developers and product people who want to ask an AI assistant plain language questions about their app's store performance, such as which countries or search terms drive the most installs.