Hi guys 👋, I’m Snow, creator of Claux. Over the last year I’ve spent thousands of hours using Claude and Claude Code for software development, automation, business operations, and product building. And while Claude itself is incredible, I kept running into the same problem: I couldn’t see what was actually happening. I couldn’t easily answer questions like: • Which projects are consuming the most tokens? • Which prompts are costing me the most money? • Which Claude Code sessions are performing best? • Why did an agent fail? • What changed between two runs? • How much am I spending this week versus last week? • What are my team members actually doing? The more I relied on Claude, the more I realized I was operating blind. So I started building this app. Claux is the interface I wished existed from day one. It sits alongside Claude and Claude Code and gives you visibility into everything happening behind the scenes: 📊 Usage Analytics Track tokens, costs, latency, sessions, skills, context, agents and performance over time. 🔍 Run History & Diff Tracking See exactly what happened between executions and understand changes instantly. 🧠 Project Memory Keep context organized across projects and sessions. 🚨 Monitoring & Alerts Get notified before costs explode or workflows fail. 📁 Project Spaces Separate clients, products, repositories, and teams. ⚡ Claude Code Insights Understand how your coding sessions evolve and where time and money are going. My goal isn’t to replace Claude. My goal is to become the operational layer around it. The same way GitHub became the operational layer around Git, I believe AI assistants need a management layer that helps users monitor, optimize, and scale their workflows. Today is my first public launch and we’d genuinely love your feedback. I am especially interested in hearing: • What frustrates you most about Claude or Claude Code today? • What visibility or analytics are you missing? • What would make your AI workflows significantly easier to manage? I’ll be here all day answering every question and collecting feedback. Thank you to everyone who takes a look. Let’s build better AI tooling together. — Snow