lcu README

The project's README from github.com/amontlabs/lcu. As one markdown file: README.md.

LCU logo

LCU

Codex computer use, decoupled from the Codex app.

https://github.com/user-attachments/assets/bd9d0809-8a6c-4931-a51c-b98daf05cb5f

LCU exposes Codex's original computer-use runtime to your harness without requiring Codex authentication. The official ChatGPT desktop app must still be installed locally: it supplies the runtime and instructions, while LCU handles setup and harness integration.

Quick start

Copy this into your agent:

Install and configure LCU for this harness using the latest release.
Follow https://raw.githubusercontent.com/amontlabs/lcu/main/docs/INSTALLATION.md
Check my OS, architecture, and prerequisites, then guide me through any required permissions.

In Claude Code on an Apple Silicon Mac, two commands do the same through its plugin manager; see Claude Code plugin:

claude plugin marketplace add amontlabs/lcu
claude plugin install lcu@lcu

Features

Approve apps from Claude, natively

The first time an agent controls an app (Zed, Notes, Safari), the computer-use runtime asks you to allow it. The Claude app's Code tab used to fail that request with "Computer Use was not approved". lcu setup --agent claude-code now installs a small Claude Code mod, lcu-approve, that shows the question as a native Computer use approval pane in the Claude app, and in the terminal.

The Computer use approval pane in Claude Code: Allow this conversation, a high-risk app with its warning, and the Approved apps panel

The /computer-use-apps command opens an Approved apps panel in the same place: one line per always-allowed app with Revoke (macOS asks for Touch ID or your password), apps that are no longer installed grouped below, this conversation's grants, and an Allow field.

Only you can answer: the model cannot see, press or fake the pane. High-risk apps show the runtime's own warning (reworded to say "computer use" and "the agent"), and the app hosting the agent is never approved. Needs the Claude app or Claude Code 2.1.287+. How it works, how to remove it.

In the Claude app, register LCU through Claude Code (lcu setup --agent claude-code), not through Settings > Connectors; see installation.

Manage approved apps with Touch ID

"Always allow" is remembered. Review and change that list from the terminal (macOS), without opening the Codex app:

lcu apps                 # what is always allowed today
lcu apps allow Zed       # Touch ID (or your password), then Zed is allowed
lcu apps revoke Zed      # take it back

The Approved apps panel opened with /computer-use-apps

Details.

Permissions

Your harness decides whether the agent may call LCU's tools (Claude's "don't ask again" on the first card, Codex's tool approval, and so on). LCU decides, per app, whether that agent may touch it, in every permission mode. lcu setup --approval auto is optional and meant for unattended machines (VMs, CI): it pre-allows LCU's two model tools in the harness, and leaves per-app approval in force. See approval mode.

Requirements

Install the official ChatGPT desktop app, Python 3.12+, and your harness first. LCU checks the installed app for compatibility and uses its original runtime. For Codex CLI, update before setup to get the required hook support.

Platform Requirements
Linux ARM64 or x86-64 Ubuntu 24.04-compatible glibc system, an active X11 desktop, and D-Bus
macOS on Apple Silicon The signed ChatGPT app, with Accessibility and screen-recording permissions approved during first use

Windows 11 x64 remains a candidate. Intel Macs, native Wayland, and musl Linux are unsupported. See verification and limits for tested behavior.

Install manually

Download the archive and matching .sha256 file from the latest release:

Platform Archive
macOS on Apple Silicon lcu-<version>-darwin-arm64.tar.gz
Linux ARM64 lcu-<version>-linux-arm64.tar.gz
Linux x86-64 lcu-<version>-linux-x64.tar.gz

Follow the installation guide to verify the checksum, extract the archive, and register LCU in your harness. Setup offers an agent chooser and desktop-readiness guidance. Chrome and computer-audio recording are opt-in.

For development, see building from source.

Enable Chrome

Ask your agent:

Enable Chrome control in my LCU setup using the installation guide:
https://raw.githubusercontent.com/amontlabs/lcu/main/docs/INSTALLATION.md#desktop-and-browser

Enable the official extension in your Chrome profile and restart your harness. Then ask it to use LCU to list Chrome tabs. Sites still require approval. Claude Code's Chrome support remains experimental because some interruptions do not trigger tab cleanup; see adapter limitations.

Enable computer-audio recording

Pass --audio when setting up a maintained harness, for example:

~/.local/share/lcu/current/bin/lcu setup --agent pi --audio

The installer does not add lcu to your PATH; run its installed path (/opt/lcu/current/bin/lcu on Linux). See the lcu command.

This enables the installed original runtime's optional computer-audio recording API and its original approval flow. LCU does not add audio-specific instructions, and a saved recording is not audio delivered to the model. See the audio opt-in verification record.

Pi commands

In interactive Pi, /lcu stop requests original Computer Use Stop for a selected app during the active LCU turn (macOS only). /lcu pick selects an original app or browser target and appends it to the editor draft while Pi is idle; review and submit the draft yourself. See adapter command behavior and limits.

Documentation

LCU is MIT licensed. Release archives contain LCU and third-party setup dependencies. The OpenAI app and its instructions come from your local installation and retain their own terms; see dependency provenance.

lcu (amontlabs/lcu)

Searching Database…

1.2K ·GitHub
Lcu
What is this?LCU extracts Codex's built-in computer-use runtime out of the ChatGPT desktop app and exposes it to any AI coding harness (Claude Code, Codex CLI, Pi), so an agent can click, type, read windows, and take screenshots on your machine without needing Codex authentication.
Does LCU replace the ChatGPT desktop app, or does it still need it?It still needs it. The official ChatGPT desktop app must stay installed locally because it actually supplies the computer-use runtime and instructions; LCU just handles setup and wires that runtime into your harness.
How do I install and set this up?Download the matching archive and its .sha256 file from the latest GitHub release for your platform (macOS Apple Silicon, Linux ARM64, or Linux x86-64), verify the checksum, extract it, and register LCU with your harness; on Claude Code you can instead install it through the plugin manager with two commands.
What do I need installed before running this?You need the official ChatGPT desktop app, Python 3.12+, and a supported harness (Pi, Codex CLI, or Claude Code) already set up; for Codex CLI you must update it first to get the required hook support.
Which platforms actually work right now?Only Linux ARM64/x86-64 with an active X11 desktop and D-Bus, and macOS on Apple Silicon with the signed ChatGPT app. Windows 11 is only a 'candidate', and Intel Macs, native Wayland, and musl Linux are unsupported.
Who actually approves an agent touching an app on my machine?Only you. Approvals happen in a native pane the model cannot see, press, or fake, and you can later revoke access per app, which on macOS asks for Touch ID or your password.
Can an agent control my Chrome tabs too?Yes, once you enable the official Chrome extension in your profile. LCU can then read and control tabs, but every site still needs separate approval, and Claude Code's Chrome support is still experimental.
Can the agent record audio from my computer?Only if you pass --audio during setup. That just turns on the underlying ChatGPT runtime's own audio-recording API and approval flow; LCU doesn't add any audio-specific instructions itself.
What can I do with LCU inside Pi specifically?Pi gets two slash commands: /lcu stop requests a Computer Use stop for a selected app (macOS only), and /lcu pick selects an app or browser target and appends it to your draft for you to review and submit.
Why build LCU instead of just using Codex's computer use directly?Codex's computer-use runtime is normally locked to the Codex app and its own login; LCU decouples that runtime so other harnesses like Claude Code or Pi can use it without a Codex account.
What are similar projects?pstack-claude, newsjack, hive, openrig, midscene, agent-desktop, awesome-chatgpt-zh, jev-browser-use
Which topics is it in?permission, pi-agent, desktop-automation, x11, codex-cli, computer-use, chrome-extension, harness, accessibility, browser-automation, chatgpt, desktop
Creator
amontlabs
Size
13.3 MB
Stars
1.2K
Updated
CommitOct 14Releasev0.11.4Oct 14
Age
26d
Commits · 90 Days