Loading video player...
Managing multiple AI coding agents across different machines and projects gets messy fast. I'd sometimes forget about a session and only recover it hours or even days later. So I built a dashboard to track all my Claude Code sessions in real-time. š Features shown in this demo: ⢠Real-time session monitoring with heartbeat detection ⢠View session metadata: device, project, git branch, model, duration ⢠Context usage tracking with token breakdown ⢠Status indicators (working, idle, awaiting permission) ⢠Remote terminal attachment via WebSocket - interact with any session from your browser ⢠Automatic stale/crashed session detection This is a prototype I built to improve my workflow as a "coding orchestrator" - providing guidance, making design decisions, and reviewing AI output while agents work on multiple tasks in parallel. š Related blog post: https://blog.marcnuri.com/boosting-developer-productivity-ai-2025