Three things, one sentence each, inside ClaudeCode/Codex: (1) all your app sales seamlessly — no more ASC dashboard, (2) daily keyword rank tracking + one-shot ASO optimization with paste-ready strings, (3) keyword-driven competitor research + review-driven feature ideation.
Hey PH 👋 — I'm an indie iOS/macOS developer (Sticky Note Pro and a few others), and AppMate started as a tool for myself.
I was doing the same things in App Store Connect every day or every week — checking sales, tweaking keywords, figuring out who was actually competing with me, reading reviews to decide what to build next. All possible on ASC's web dashboard. None of it fast.
AppMate is autopilot for iOS app growth, inside ClaudeCode/Codex. Three things, one sentence each:
**[1] All your app sales × ClaudeCode/Codex, seamless.**
Never open the App Store Connect dashboard again. Your AI CLI always has direct access to all your app data — sales, downloads, 5 time windows (D-1 / 7d / 30d / WTD / MTD), all live apps, auto-anchored to Apple's 1-2 day reporting lag.
**[2] One-shot ASO optimization, daily keyword rank tracking.**
Every day, AppMate auto-tracks keyword ranks across your apps (iTunes Search Top-200 lookup, diffed against yesterday's snapshot). When you want to optimize, one sentence produces paste-ready strings (title 30 chars / subtitle 30 / keywords 100) with OLD vs NEW comparison and a plain-English reason for every add and delete.
**[3] Keyword-driven competitors → review-driven feature ideation.**
First, find the rivals actually stealing your traffic — not generic "similar apps," but the ones outranking you on your own core keywords. Then combine those rivals with your last 90 days of raw reviews (LLM classifies into complaint / suggestion / praise) to tell you what to build next, two sentences per idea.
Two things I care about beyond features:
· **Read-only by enforcement.** AppMate refuses to run with any App Store Connect API key that has write capability — runtime role probe + code-level write block. Even an over-privileged key can't damage your live app data.
· **Same skills run across ClaudeCode/Codex** from a single source.
Design pattern: Python does deterministic data work (API calls, caching, rank lookups). The LLM handles what needs judgment — CJK/Latin tokenization, candidate generation, strategy reasoning, report rendering in whichever language you're conversing in.
Free, open source, no telemetry. Happy to answer questions.
GitHub: https://github.com/fengyiqicoder...
Report
No reviews yetBe the first to leave a review for AppMate
BoxAi