Coordinate parallel coding-agent worktrees and terminals with Ouijit
Use Ouijit to manage parallel CLI coding-agent tasks with per-task git worktrees, integrated terminals, lifecycle hooks, and sandboxed sessions.
npx skills add agentskillexchange/skills --skill coordinate-parallel-coding-agent-worktrees-and-terminals-with-ouijit
Use Ouijit when an operator needs to supervise several local coding-agent workstreams without manually juggling branches, terminals, task state, and review handoffs. The workflow is to add a git project, create task cards, let Ouijit create per-task worktrees, launch Claude Code, Codex, Pi, or OpenCode from task terminals, and use hooks or the session-aware CLI to start, continue, review, or complete work.
What this skill actually does
This is skill-shaped because the repeatable job is bounded: coordinate local agent coding tasks through isolated worktrees, terminal sessions, lifecycle hooks, and optional VM sandboxing. Invoke it when the operator wants a controlled multi-agent cockpit around existing CLI agents, not as a generic Electron app, task board, terminal emulator, or replacement for the underlying agent runtimes.
Inputs and prerequisites: Ouijit desktop app or source build, Git, Node.js 20+ for source builds, Claude Code/Codex/Pi/OpenCode or another supported agent CLI, optional Lima VM or nono sandbox backend.
Setup notes: Download the latest macOS or Linux release from https://github.com/ouijit/ouijit/releases/latest, or build from source with git clone https://github.com/ouijit/ouijit.git && cd ouijit && npm install && npm start. Add a git project, create task cards, configure lifecycle hooks for the chosen agent CLI, and use Ouijit’s terminals or the ouijit CLI to create, start, continue, review, and complete tasks.
Source and verification boundary: use https://ouijit.com/docs/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.