The terminal AI agents can drive — and the cockpit to run them from. 99 MCP methods / 21 namespaces, agent Inbox, verified worktree fleets, diff review & rollback. macOS / Linux / Windows, local-first, $0, MIT.
This repository also configures its own agents. See what unterm tells them →
Skill Claude CodeCodex
Orchestrate multiple AI coding agents in parallel terminal panes — fan work out (e.g. "fix this bug 3 ways", "have Claude/Codex/Gemini each take a stab"), watch their progress, and reconcile results. Uses Unterm MCP to spawn a pane per agent, drive each one, and aggregate. Activates when the user asks to "run agents…
Skill Claude CodeCodex
Cut a release for the current project end-to-end — version bump, tests, tag, push, build artifacts, upload — driving the user's real terminal via Unterm MCP. Activates on "release vX.Y", "ship a new version", "cut a tag", or "publish the release". Requires Unterm MCP available; otherwise decline and suggest…
Skill Claude CodeCodex
Drive a real macOS / Linux / Windows terminal from this conversation — spawn shells, split panes, type commands, read scrollback. Activates whenever the user wants you to "run X", "open a shell and ...", "split the pane", or "look at what's on the terminal", and the Unterm MCP server is available.