spawning-a-session

To spawn a new Claude Code session, you provide a baked-in first prompt, select a git worktree, and assign a remote-control name. The Skill enforces pre-flight artifact checks and records a post-spawn status file so the agent can be managed later.

Yes, autonomous Claude Code sessions can be launched across iTerm2 panes, tabs, and windows, as well as shell, tmux, SSH, and headless environments. A nohup fallback is used when no terminal integration is available.

Orchestrating multiple Claude Code agents for sprint kickoffs and hotfixes is best handled by spawning self-contained sessions with persistent remote-control identities. This enforces permission skipping and post-spawn status recording for reliable delegation.

Yes, spawning a Claude session requires pre-flight artifact availability, worktree selection, session naming, and autonomous permission configuration. These prerequisites ensure the spawned session operates reliably and can be tracked via a status board.

A spawned Claude Code session needs a persistent remote-control identity so orchestrators can find, manage, and delegate work to the agent reliably across long-running projects. It also enforces post-spawn status recording for tracking.

The nohup fallback for spawning headless Claude sessions is limited by the absence of terminal integration, meaning it cannot leverage iTerm2 panes, tabs, or windows. It serves as a fallback when standard shell or tmux environments are unavailable.