Project kickoff

UserActivity.ai is officially in motion.

The target audience isn't enterprise product teams — it's vibecoders. People building apps with AI coding tools like Cursor, Bolt, and Claude Code who ship fast but fly blind. They can generate a full-stack app in an afternoon, but they have no idea if anyone's actually using it or where users get stuck.

Traditional analytics tools don't work for this audience. They're complex to set up, require manual event instrumentation, and surface data in dashboards designed for analysts. UserActivity.ai takes a different approach: one script tag in your HTML, and you're done. The system watches user behavior automatically and surfaces insights in natural language — not charts and funnels.

The architecture is a pnpm monorepo with turborepo. The marketing site is Next.js 15. But the real product surfaces are AI-native: a CLI tool that lets you ask questions about your users in plain English, and an MCP server that integrates directly into AI coding agents. Your coding assistant doesn't just write code — it can see how users interact with what it built.

That's the core thesis: AI coding agents are building the software, but they're doing it without any sense of how that software actually gets used. UserActivity.ai gives them behavioral sight. The tagline is "the open-source behavioral intelligence layer for AI coding agents."

The domain is on Cloudflare (useractivity.ai), and there's a GitHub org (UserActivityAI) for the public-facing repos.