zhitongblog

Unterm — AI skill for Claude Code

AI community

The terminal AI agents can drive — and the cockpit to run them from.

How to install Unterm

This entry records only its repository, not the path inside it, so there is no exact command to give. Open zhitongblog/unterm and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Unterm does

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.

Alternatives in AI

Open Science — Open Science Desktop — local-first, model-agnostic AI research workbench for macOS, Windows & Linux 1.5k ★
OpenGhost — Open desktop AI agent for Windows, macOS and Linux, written from scratch: its own agent engine, its own browse 266 ★
Hazmat — Containment for AI agents - user isolation, sandboxed execution, network controls, backup/rollback 167 ★

README

Unterm

**The terminal AI agents can drive.**

![Agent Cockpit: three agents working, the Inbox surfaces the one that needs you, Enter jumps to it](assets/demo/agent-cockpit.gif)

Cross-platform terminal (macOS / Linux / Windows) built on Unterm's native `next-core` terminal engine, with one design bet: the terminal itself is controllable from the outside by any AI agent over MCP. Claude Code, Codex, Gemini CLI, Cursor, Aider, your own scripts — they all get the same JSON-RPC surface (**149 authenticated methods plus `auth.login`**) to spawn shells, run commands, read pane state, capture screenshots, change settings, and record sessions.

Since v0.68 the terminal is also something an orchestrator can govern rather than merely call: it publishes what it can do and how dangerous each capability is, works under leases that expire and cannot be replayed, keeps agents inside workspaces that cannot see each other, and can hand you an evidence bundle for a task that somebody who was not there can verify.

Since v0.55 the relationship runs both ways: agents drive the terminal from outside, and the terminal is an **Agent Cockpit** for the agents running inside it — live per-pane agent state, a waiting-first Inbox, fleets of N agents on one task in N isolated git worktrees, and a Review page to diff / merge / roll back what they produced.

The other 2026 terminals each pick a different side: Warp embeds AI inside a closed cloud (Oz), Ghostty stays out of your way and lets you bring your own tools, iTerm2 is Mac-only. Unterm picks the third side — terminal as MCP-controllable surface, deliberately keep AI *generation* out of the terminal, let external agents grip it through the API, and give the human one cockpit to run them all from.

Practical implications:

Every Unterm window starts a local MCP server (line-delimited JSON-RPC over TCP) and a local HTTP settings server (Web Settings page) on auto-allocated ports. Both are auth-token gated, both are 127.0.0.1-onl