Hi everyone! Hope you had a great day so far, and maybe its about to get just a little bit better (thanks Winter ;) So I had way to many terminal windows flying about when using Cl…

episko gives every Claude Code session its own terminal, watches what each one is doing, and gathers them into a single view — so a whole flock of agents is as easy to mind as one chat. Each session is one little Claude, coloured by what it's up to — amber is busy, green is done and waiting on you, pink needs a decision. Every Claude runs in its own live terminal — here, or in Ghostty, Terminal, or iTerm. Type to it, watch it think. It's the real process, not a replay. Model, context, cost, and the exact tool it's running right now — read live from each session, with a little history of what it touched. When an agent needs a yes or no, episko brings the question to you — the command, and how risky it looks. Allow, deny, or hop into the terminal. ⌘K opens one little search over everything — jump to any Claude, start a new one, or run a command. Whoever needs you floats to the top. Point a new Claude at the repo, one of its worktrees, a branch, or a fresh branch you name — and see where it'll land before you send it. The scripts your project already ships — package.json, a justfile, VS Code tasks, a Makefile — run in these panes too. A run is just another session: its exit code is its status.