v0.3.3 now available

Every coding agent.
One calm inbox.

Shipwright Lab turns your terminal into a vertical inbox of projects and surfaces the local coding agents that need your attention. Agents waiting on input rise to the top; everything else stays quiet.

Download for macOS View on GitHub

Signed & notarized · Requires macOS 13+ · Apple Silicon + Intel

Shipwright Lab showing a project inbox of coding agents beside a live Claude Code session

Drives the local agents you already run

Claude Code Codex Other local CLI agents No account required

Parallel session management

Run a dozen agents at once, never lose the thread

Kick off as many agents as you want across every project and let them work in parallel. Shipwright watches each session and pulls the ones that need a decision to the top with a status dot. The moment an agent asks a question or finishes a run, you know. When every agent is heads-down, it simply reads "Nothing needs attention."

Track every running agent in one place, grouped by project
Live Asking / Running / Done states, sorted by who's waiting on you
Unread badges and input-needed alerts so nothing finishes unnoticed
Quiet by default, no noise when nothing needs you

Shipwright Lab project inbox sidebar with agents grouped by project and status dots

Local agents, no cloud

Your machine. Your agents. Your keys.

Shipwright Lab drives Claude Code, Codex, and other local CLI agents through local hooks (OSC 777). Nothing is brokered through a server; there's no account to create and no session leaves your Mac.

Connects via standard terminal hooks (OSC 777)
Bring your own CLI agents and credentials
No cloud relay, no telemetry, no sign-in

Shipwright Lab running a Codex agent session next to the project inbox

Resume sessions

Pick up exactly where you left off

Close a tab, quit the app, or come back tomorrow, and your agent conversations are still there. Shipwright reliably reattaches to each Claude Code and Codex session, so one click drops you back into the full context with history intact. Resume a single tab, or bring an entire project's agents back online at once.

Resume per tab or restore every session in a project together
Full conversation and scrollback restored, not just a fresh prompt
Survives app restarts; jump back in days later without losing the thread

Shipwright Lab resuming a saved agent session with full conversation history

Built for long runs

Details that keep you in flow

Built-in file explorer

Browse the project tree beside the active agent: open a file, check the structure, and keep the conversation going without leaving the window.

Search your sessions

Find that command, error, or answer from any past run. Search across every agent session, not just the scrollback in front of you.

History & autocomplete

Built-in command history and inline autocomplete: recall what you ran before and complete commands as you type, without leaving the agent.

Built-in markdown viewer

Plans, specs, and READMEs render right inside the app, so you read what your agent writes as formatted markdown, not raw text.

Code review panel

Review every change your agent makes in a built-in diff panel: see exactly what was edited before you accept it.

Sail keep-awake

One click keeps your Mac awake and the screen on, so long agent runs finish unattended instead of stalling on sleep.

Fast native terminal

Built on a GPU-accelerated core, so scrollback and live agent output stay smooth even under heavy load.

Light & dark, done right

A graceful, modern interface in any light

Shipwright is designed to feel calm and considered the moment you open it. Switch between a deep, focused dark mode and a clean, airy light mode, and the whole interface follows, from chrome to agent output.