claude fleet.
the problem.
in this era, when ai coding is this much faster than manual coding, you can either choose to take more free time or run more agents in parallel. i chose parallel.
the real skill comes from context switching between multiple ai coding sessions, but you can build tools to make it easier. that's what led me to make claude fleet - my dashboard for ai agents.
i took inspiration from discord overlay - a semi-transparent, always-on-top window that shows what every claude session is doing across my machine in real-time.
the overlay sits on top of everything like discord's overlay, showing compact session pills that update in real-time.
how it works.
the app hooks into claude code's webhook system to track session lifecycle, tool usage, and permission requests. the tricky part: when a user cancels an agent (pressing esc), no hook fires. so i added file watching on claude's transcript files to catch those edge cases by parsing the jsonl for interruption patterns.
tech stack.
- rust + tauri 2.0 — native desktop
- next.js 15 + react + typescript — frontend
- tokio async runtime + axum — hook server
- notify crate — cross-platform file watching
- websocket broadcast — sub-50ms ui updates
the app works by running three local servers: http hooks (14322), websocket broadcast (14324), and an mcp server (14325) for direct claude integration.
what it tracks.
- session status (coding, waiting, needs help, done)
- editor detection (vscode, cursor, neovim, windsurf, continue, claude cli)
- git branch and repo info
- files changed, current task
- process/terminal tracking
future.
- support for models outside of claude code
- one-click navigation to the repo when a session finishes, taking you straight into the repo you're working in instead of looking through countless repos