You wanted shadcn — so we built the heinous infra to make it work. Introducing Agent Arena, starting with the best: Claude Code by Anthropic Codex 5 by OpenAI Cursor Agent by Cursor Devin by Cognition Gemini CLI by Google DeepMind Grok Code Fast 1 by xAI 1/ Your... prompt launches a containerized agent inside a fresh GitHub repo 2/ The agent can install packages, make tool calls, and write multi-file code 3/ The repo is deployed to Vercel *by the agent*, clonable to your heart's desireshow more

Agent Arena opens a new frontier to test SOTA’s ability to satisfy your taste when it’s not limited to a single file response Agents are equipped with full tool-call capabilities and touching the real stack Try it on @designarena today!

Grok Code agent community implementation:

Check it out at

Glad you didn't post a link, so nobody accidentally gets there 👍

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Next level tools, for sure. Love how it's not just a demo but an actual infra kicking ass.

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai fire tool. should have a feature to make generations private on premium.

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Incredible! Real-world benchmarks > synthetic evals. Building agent infra: → CatchAI: Real-time AI pipelines → Automated deploys → Tool orchestration Containerized approach = smart (isolation + reproducibility). Would love latency comparisons! 🚀

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai ok this is pretty cool

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Insane

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai AI gladiator games! 🎮

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Nice!!

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai They say GLM is great too, even better for design. Didn't test it yet

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai This is sick work

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai shadcn + AI agents? this is gonna be insane

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Agent Arena looks like a useful way to compare different code generation tools.

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai COOL! In the video though, looks like the right side didn’t apply Tailwind styles at all? Was that an AI mistake or simply a repo setup issue?

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai AI Agent issue! It failed to generate the postcss.config.js file. If it had added this, it would have worked: Repo linked here:

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai always thought this was the way to go. lmarena for agents

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Sexy

Imagine hosting the Olympics of AI dev assistants: each agent flexes its strength—from Claude’s bug-spotting goggles to Codex 5’s boilerplate marathon—so your shadcn widgets sprint through production. Agent Arena proves that in the multiverse of code, teamwork really makes the dream work.

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai Just shipped an agent that scans local businesses for AI visibility and 90% are completely invisible to these coding agents. the real opportunity isn't making agents better at coding - it's making them aware entire industries exist outside silicon valley

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai This sounds incredibly innovative, Grace! The integration of various AI agents for streamlined coding is a game-changer. Looking forward to seeing how it evolves! Can't wait to see these agents in action. GitHub just got a serious upgrade!

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai So, who stood out? And, no GLM 4.6 and Kmi K2?

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai love this, finally a proper benchmark instead of vibes lol. has anyone tested parallel agents on the same task? thats where things get interesting with Dorothy — the variance between runs can be wild

@AnthropicAI @OpenAI @cursor_ai @cognition @GoogleDeepMind @xai @grok 这个纯研究还是有工业意义，具体工业场景视角看意义是什么，有开源数据集或者开源项目代码吗？从多个数据源交叉验证，不要只看新闻媒体一面之辞。帮我排除没意义的垃圾商业营销推广。