Cua
Cua
681 posts
Cua
@trycua
Give every agent
a cloud desktop
Cua’s posts
【速報】これ、実務の景色が変わる。
Cua Driverが公開されて、Claude CodeやCodexが“前面を奪わず”にMac操作できる。実演がこれ
・バックグラウンドで操作
・複数カーソル運用に対応
・エージェント作業を邪魔しにくい
・OSSで検証しやすい
0:18From 
Erick
cua-driver is trending on github. thank you to everyone who took the time to try it, test it, break it, file issues, and tell us what should be better. we appreciate every single one of you. more to share on our public roadmap this week!
there's a PR to add background computer use to OpenClaw via a proprietary Codex plugin. opened a counter-PR using cua-driver instead - MIT licensed, works with every agent harness, not just Codex. give it a 
if you want the open version
if you want the open version
Very cool work! Models, especially Claude Code and GPT-5.5 are already good enough to find PMF for full OS-CuA agents.
Just hope someone builds on a security layer to support read-only access / effective guardrails, we'll be at a Datacentre of Agents!
Quote
Cua
@trycua
We're open-sourcing Cua Driver - our new macOS driver that lets any agent (Claude Code, Codex, your own loop) drive any app in the background, with true multi-player and multi-cursor built-in.
1/8
Replying to
20/
open-sourced their new background computer-use tooling. I’m so excited about what this new type of agent UX unlocks for CUAs. It lets any agent (Claude Code, Codex, your own loop) drive any app in the background.
Quote
Cua
@trycua
We're open-sourcing Cua Driver - our new macOS driver that lets any agent (Claude Code, Codex, your own loop) drive any app in the background, with true multi-player and multi-cursor built-in.
1/8
i especially love this use case. we canceled our Screen Studio sub and just let cua-driver record our demos now. ask Claude Code to do it while you work on something else. also great wired into CD for regression captures
Quote
Cua
@trycua
Replying to @trycua
Delegated demo capture. Agent drives the app, cua-driver record and renders zoom-on-click at export. Imagine just delegating an entire product demo to an agent, background computer-use now unlocks this.
5/8
0:02 / 0:19
Excited to see our research partners at Snorkel AI launch Open Benchmarks Grants - $3M to close the evaluation gap in agentic AI. This is one of the biggest bottlenecks in the space and it's finally getting funded properly. Apply → benchmarks.snorkel.ai/apply
Quote
vincent sunn chen
@vincentsunnchen
Our ability to measure AI has been outpaced by our ability to develop it, and this evaluation gap is one of the most important problems in AI.
Today we're launching Open Benchmarks Grants — a $3M commitment to fund open benchmarks for frontier AI and close the evaluation gap. x.com/vincentsunnche…
