Authorization notice. Lcu is dual-use. Only use it on software, sites and systems you own or have explicit permission to test.
Remote script. Setup runs a downloaded installer and unattended approval lets the agent act without a person present, so use it only on disposable machines.
Who it is forA DevOps engineer or tester who runs agent-driven GUI checks on a VM or CI runner where no one can answer approval prompts.
The problemInteractive approval prompts block unattended jobs, so GUI test runs stall when no one is present.
The outcomeThe agent runs GUI steps on the disposable machine without prompts, and per-app approval still applies to every app it touches.
What you need first
- The official ChatGPT desktop app installed locally
- Your agent harness installed (Claude Code, Codex CLI, Pi, or another supported adapter)
- macOS on Apple Silicon, or Linux ARM64 or x86-64 with an active X11 desktop and D-Bus
- Accessibility and screen-recording permissions on macOS, approved during first use
Steps
- Install the official ChatGPT desktop app and your harness on the disposable VM or runner.
- Run
~/.local/share/lcu/current/bin/lcu setup --approval auto. (inferred, check the README) - Run
lcu appsto confirm which apps are always allowed on this machine. - Run the agent job, then review the results and any screenshots it saved.
- Run
lcu cross-turn offif you turned on cross-turn computer use for this job.
Prompt for your agent
Using the amontlabs/lcu repo, configure LCU for unattended use on this disposable machine with `lcu setup --approval auto`. Run `lcu apps` and report the list of always-allowed apps. Verify that the list contains only apps the job needs.
Caveat. The auto approval mode only pre-allows LCU's two model tools in the harness, and per-app approval stays in force.
Get 5 new use cases every morning. Free, one short email, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Disclaimer
Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page