Mekari

So I've been digging into the Android Vitals numbers for my android app lately, and the process was kind of painful. Every time something looked off — slow startup, high ANR rate, janky rendering — I'd open Play Console, filter by version, switch to a different tab, try to figure out which build caused it, which API level was affected... and by the time I had a clear picture, I'd already lost 30 minutes just gathering context. The data was always there. I just had to hunt for it manually every single time. So I built something to fix that. android-vitals-mcp connects your AI coding assistant directly to the Google Play Developer Reporting API. Instead of clicking through dashboards, you just ask: "why is the cold start getting slower?" "which version caused the ANR spike?" "the app feels janky on Android 14, what happened?" And your agent actually goes and checks — pulls the real numbers broken down by version code and API level, compares them against Google's thresholds, and walks you through what to fix. Setup takes two commands: npx android-vitals-mcp setup npx android-vitals-install-skills The first one is a wizard that connects your Google Service Account and writes the MCP config into every AI CLI it detects. The second installs four skill playbooks that auto-trigger on keywords like "ANR", "jank", or "cold start" — so you don't even have to think about it. It just kicks in when it's relevant. Then restart your AI CLI and ask: "improve my Android crash rate". Works with Claude Code, Gemini CLI, Cursor, Copilot, Codex, Kiro, and more. If you're spending more time reading Play Console than actually fixing things, give it a try😬 📦 https://lnkd.in/gpYzJnri #Android #AndroidDev #MCP #ModelContextProtocol #AI #AgentAI #ClaudeCode #GooglePlay #AndroidVitals #MobilePerformance #MobileEngineering

Over the past month, I’ve been building my own SEO automation system, because managing SEO manually at scale is no longer realistic. When you’re dealing with thousands, tens of thousands, even hundreds of thousands of pages, the traditional approach starts to break down. Auditing content one by one, manually checking rankings or indexing, and optimizing page-by-page simply doesn’t scale. So instead of focusing only on optimizing pages, I shifted toward optimizing the system behind them. I’m developing automation that supports key SEO workflows from data extraction and clustering to content auditing, internal linking, and optimization recommendations. The intention isn’t to replace SEO work entirely, but to meaningfully reduce manual effort, covering around 20–30% of operational tasks, so the team can focus on higher-impact strategy and decision-making. In high-scale environments, success comes not from working harder on each individual page, but from building systems that enable every page to perform at its best.