I built App Store Operator because I live in Claude Code and I kept breaking flow to do App Store work. Researching a keyword meant opening the App Store, finding the top apps, switching to SensorTower, reading download and revenue estimates, copying them into a doc — 15 to 20 minutes per keyword, every time. Writing In-App Event copy meant a separate context switch into a character-counting spreadsheet. None of that output landed inside Claude where I was already thinking and building. So I built an MCP server that brings it all in. Once it's set up — one terminal command — you just ask Claude in plain English. It handles the rest. No tabs. No tools. Just prompts. Here's what it does today: 1. research_rivals — searches the App Store for any keyword, scrapes SensorTower for the top 3 results, and returns 16 structured fields per app: downloads, revenue, ratings, publisher, top markets, and more 2. search_app_store — ranked App Store results as a structured table, up to 25 apps, no browser required 3. get_app_details — on-demand SensorTower analytics for specific app IDs when you already know who you're watching 4. prepare_iae — generates iOS In-App Event copy in 3 variations with all App Store Connect character limits enforced, in any of 14 supported locales, no external API calls What's coming next is the part I'm most excited about: full ASO through the App Store Connect CLI — updating metadata, submitting copy, managing your entire App Store presence — without ever leaving Claude. The goal is a complete ASO workflow that lives entirely inside your AI conversation. It's MIT licensed, published on npm, and free. A 24-hour local cache means repeat queries on the same keyword return instantly. I'd love to hear from other developers living in Claude Code — what App Store workflows are still costing you the most time? What would make this genuinely more useful for how you actually work? Thanks to everyone in the Product Hunt community for taking a look today.