We released open-source apps written entirely by AI, including a firewall and an OS. Tear them apart.

QuickOpen publishes software that is 100% AI-written: a server firewall, a VPN client, a mail client, a browser built on ungoogled-chromium, an office suite, LocalLLM Studio and a Linux-based Quick OS. Source and releases are on the site.

The obvious question is whether any of it is safe to run. The trust page documents the build, test and review process, and I'd rather you audit it than take our word.

What would you want to see before running an AI-written firewall on a real server? https://quickopen.ai/browse