Jev is now available to everyone. No waitlist. Start using it here: console.typesafe.ai

7 UI skills every design engineer should know when building with AI agents: 1. emil-design-eng ui-skills.com/skills/emilkow… 2. make-interfaces-feel-better ui-skills.com/skills/jakubkr… 3. playwright-cli ui-skills.com/skills/microso… 4. react-doctor ui-skills.com/skills/million… 5. fixing-accessibility ui-skills.com/skills/ibelick… 6. 12-principles-of-animation ui-skills.com/skills/raphael… 7. shadcn ui-skills.com/skills/shadcn-…

Introducing the ElevenLabs MCP, now available in Claude. Your team can manage your voice and chat agents where you already work - review recent performance, create new agents, update configurations, and even estimate LLM costs before changes go live.

Astra can create beautiful three.js UIs entirely in code. You don't need an assets library or MCP to create detailed 3D models. These are all code and are easy to customize for your agent. Give Astra a video, URL, or code reference, then ask it to keep comparing the result against the reference and fixing mistakes. Liquid Glass, shaders, motion design, procedural three.js models, and full environments. All from a few prompts. Some prompts I used: "Recreate this in a single HTML file with three.js. Keep comparing it against the reference and fix any differences." "Add Liquid Glass buttons. On hover, animate the lighting around the borders and make the background gradients move slightly with the pointer. Add subtle particles." "Improve the lighting, edges, and modeling. Give the keys more depth and verify the result against the real MacBook." Too many projects I could open-source. Let me know which one should I release.

We're open-sourcing Claude Commerce Agents. This is a blueprint for building shopping and merchant agents, with reference implementations across retail, travel, telecom, and entertainment.

I haven't released a new course in two years because I thought AI would replace education. I was wrong. In a sea of AI slop, good education has never felt more relevant. So I'm releasing DesignCode 5 with a new Codex masterclass. The new site was built entirely in Codex with lots of three.js components: designcode.io/ I added homework, tests, certificates, a vault, and weekly live sessions. Pro and Lifetime members also get the Product Pass, with one year of Pro access to Framer, Mobbin, Aura, and more. GPT-5.6 Sol xHigh migrated ten years of content from a large CMS and Dropbox into Firebase. That made me realize I didn't need to build another CMS. With the right skills, Codex can handle video editing, writing, image generation, screen recording, and content updates. I'm spending less time managing everything and more time creating. It's never been more fun. I created a bunch of skills for this workflow. They're all in my repo, and the Three.js components are on threeui.com. Huge thanks to @framer and @mobbin for believing in me and giving our Pro and Lifetime students one year of Pro access. To every new student, welcome. I'll be there every week, giving you direct feedback and teaching you everything I know. DesignCode 5 is 50% off for early adopters right now. There's also a $499 Lifetime option for anyone tired of another subscription.

We built vgpu.sh to ship shaders on vercel.com. Now it's open source. ▪︎ Minimal agent-first WebGPU library ▪︎ Run in the browser or headless Node.js ▪︎ Render in CPU sandboxes and CI tests ▪︎ Create reusable .wgsl modules

The WebMCP Challenge is here. We’ve teamed up with @ChromiumDev, @CloudflareDev, @ShopifyDevs, @vercel, @render, and @Netlify for a 10-day hackathon. Up for grabs: $35,000 in cash prizes, Codex Micros, ChatGPT Pro subscriptions, and more prizes from our supporters.

🚨 ÚLTIMAS NOTÍCIAS! O engenheiro que criou o Claude Code do zero acabou de publicar um vídeo de 28 minutos que é ouro puro : como escrever prompts que realmente entregam resultados absurdamente bons. Já vi cursos de US$ 300 que nem chegam perto do que ele explica nos primeiros 10 minutos. Arquivos CLAUDE. md, atalhos de memória, sessões paralelas e padrões de prompting que quase ninguém usa… Tudo em um único vídeo. Totalmente gratuito. Sem enrolação. Não importa se você é desenvolvedor, está começando agora ou já usa o Claude há meses : isso muda o jogo para você a partir de hoje. Salve. Assista. Compartilhe.

I open-sourced ThreeUI, my library of three.js components and landing pages. 160+ are free, and the tool is free too. It includes procedural 3D hero sections, icons, and motion designs. Copy the prompt or source, give it to your agent, then change the theme, lighting, motion or layout. Live site: threeui.com Repo: github.com/MengTo/threeui I'm adding a lot more soon. Pro comes with 50+ extra components, MCP, and skills. Early adopters get 50% off right now.

Claude Academy is now live. Whether you're figuring out what AI is or already using Claude every day, there's a path that meets you where you are. The courses and tutorials are free and open to anyone at academy.claude.com

syncing to and from @github is now supported in ai studio build start from an existing repo, push and pull changes, and seamlessly work across environments ai.studio/build to get started

Sol, Terra, and Luna, our GPT‑5.6 family of models, are starting to roll out now in ChatGPT, Codex, and the API.

I'm open-sourcing my Agent Skills library. 75 skills for Codex, Claude Code, Cursor, and other agents, focused on web design, landing pages, motion, WebGL, UI styles, and assets. A few favorites: - Video to Super Prompt Turns a screen recording of a design, landing page, or animation into a super detailed prompt that Fable 5 can one-shot into HTML. - HTML to Interaction Prompts Takes an existing HTML page, like something built in Aura, and extracts prompts for sections, buttons, animations, WebGL effects, and interactions. - Stitched Full Page Capture Captures the entire landing page, not just the hero, so you can use the full page as a design reference. - Daily UI Inspiration Combines multiple skills into an agent loop that browses the web, captures great landing pages, and turns them into detailed prompt packs. It's free. Fork them and adapt for your own workflow.

shipping fast vs shipping smart There's this weird thing happening in my founder friend group lately where everyone's obsessed with shipping velocity. Like, genuinely obsessed. I'll be on a call with someone and they'll be like "Yeah we shipped 12 features this month" and there's this pride in their voice, like they just told me they ran a marathon. But then I ask "Cool, which ones are people actually using?" and there's this pause. And I get it. The AI tools make it sooooo easy to ship now. You can go from idea to live feature in like 6 hours with Cursor, Bolt, Lovable, V0 and Claude. The dopamine hit of seeing something you built in production is addictive. I did the same thing when I first discovered no-code tools in 2019. I was building like 3 apps a week just because I could. Felt like I was crushing it. But through that I learned there's shipping to learn something and shipping because you feel like you should be shipping. The difference is huge. Shipping to learn looks like: "We think our users want XYZ, let's build the simplest version and see if they actually use it." Shipping because you can looks like: "We haven't shipped anything in a week, let's add that feature we talked about." When you're optimizing for output instead of outcome thats a trap. You ship five features in a month and wonder why their metrics didn't move. The ability to build fast became a drug, and now a trap is chasing the wrong high. The teams that are actually winning have one thing in common They document everything. Not for users, for themselves. What they're shipping, why they're shipping it, what they expect to learn. sometimes thats in words, sometimes that's in founder-led video. This creates this weird insight flywheel. Writing forces you to articulate your assumptions. Those assumptions become testable. The results feed back into better decisions. The act of writing "we're building this because we think X" makes you realize how often you don't actually know what X is. The friction to ship got so low that the friction to think feels relatively massive. It's easier to just build something than to sit with uncertainty about whether it's the right thing. But the best builders I know have gotten really good at sitting with that uncertainty. They'll spend days just talking to users before writing a single line of code. They ship fast when they need to learn something. They ship slow when they already know what they're building. So if you're feeling the pressure to ship constantly Ask yourself: am I shipping to learn or am I shipping to feel productive? Both are fine. Just know which one you're doing.

⚡ INTRODUCING: T3MP3ST!!! ⚡ AUTONOMOUS HACKBOT STRIKE FORCE 🌩️ BRING THE STORM 🌩️ your favorite coding agent is now a full-stack red team 🫡⚔️ github.com/elder-plinius/T3M… that AI agent already humming in your terminal? well now it has FANGS. strap a full offensive-security harness onto the agents you already pay for — Claude Code, Codex, Hermes, etc. — point it at an authorized target, and in a few clicks you're watching it hunt real vulns autonomously! T3MP3ST is a harness of harnesses, with prompting that unlocks offensive-cyber workflows + a full arsenal of exploit tooling that'd make any seasoned hacker smirk. simple, yet powerful. 🦾 support for: 🕸️ web apps, APIs, OWASP Top 10 🔌 network recon + fingerprinting (live nmap/DNS/HTTP); lateral + privesc experimental 📂 source code audits, white-box vuln hunting 🚩 CTFs, wargames, challenge ranges 💰 smart contracts / DeFi / Solidity (reproduction — Damn Vulnerable DeFi, not novel discovery) 🤖 embedded, IoT, OT/SCADA, robotics OSS … and more in development! now let's talk numbers 👇 📊 XBEN — XBOW's own 104-challenge suite: • black-box: 90.1% pass@1 from the single-agent exploit loop (worst single sweep 91/104 = 87.5%) — clearing XBOW's past self-reported 85% on their own suite. gpt-5.5. • white-box (source staged, reported separately): 98.7% pass@1, worst single sweep 102/104 = 98.1%. 🎯 every solved flag graded reported-vs-expected against the challenge's own committed flag oracle — `verify-claims` recomputes the pass/fail from committed artifacts. looks like we need new benchmarks 😏 🧩 Cybench — the 40-task academic bench (Opus 4.8, hints + writeups stripped): 23/40 = 58% single-run, hint-free pass@1 — real exploits (format-string pwn, eval-jail escapes, crypto oracles), every flag graded vs a committed oracle. (Anthropic reports 76.5% pass@10) 🕳️ CVE-Zero — we pointed it COLD at real CVEs disclosed in 2026, AFTER the model's training cutoff: 10 unseen 2026 CVEs across 7 languages — prompts never tuned on them. a single agent pinned 8/10 to exact file/line/CWE (stable under re-scoring); the full pack surfaced all 10. memorization AND overfitting, both off the table — it's finding real vulns whose disclosures landed AFTER the model's training cutoff. (n=10, reported honest & directional) 🧠 the architecture: either run as a SINGLE agent (already the benchmarked, incredibly-capable path) — or pack-hunt with dozens of agents running on 8 specialist operator classes keyed to Cyber Kill Chain + MITRE ATT&CK phases: recon → scan → exploit → lateral → exfil → persistence → C2 → report. ⚓️an Op Admiral plans the whole op from a plain-english target. flip on coordination (experimental) and the operators share a blackboard — a tool-verified finding spawns the next move. full swarm or solo one operator, your call. the admiral can also update the prompts, tools, and configs of the other agents on the fly, and T3MP3ST gets stronger the more memories you build! 🧰 the Arsenal is comprehensive — nmap / nuclei / semgrep / ffuf / gobuster + more. 35 wired by default (the clean bench runs bash-only for a comparable number), 83 with the opt-in full arsenal (T3MP3ST_FULL_ARSENAL), and the spicy post-ex drivers (metasploit, hydra) gated behind human approval. exposed via CLI + HTTP API; recon (security_recon) is also live over MCP so your agent invokes it natively. 🔗 🛰️ where this goes: a self-improving swarm of specialist operators wielding a full Kali+ arsenal, learning which loadouts + configs are the most efficient tactics available, WITH a held-out train/test split baked in so it can never fool itself on its own eval. built in the open, one re-derivable number at a time. 🚧 this is v1, and parts are still under active development. chunks of the arsenal, the coordinated swarm, and some ranges are still being wired up. it's built in the open, and the receipts tell you exactly what's live vs what's roadmap. offensive security shouldn't be pay-to-play. T3MP3ST puts a red team in the hands of anyone with a coding agent. what's the first target you're feeding it? 👇 ⚠️ DISCLAIMER: FOR AUTHORIZED USE ONLY. point it only at systems you own or have explicit written permission to test. unauthorized access can be a crime, and that call is yours alone. shipped as-is under AGPL-3.0: no warranty, no liability, zero endorsement of misuse. get permission. stay in scope. open source. AGPL-3.0. 100% free. FORTES FORTUNA IUVAT 🌩️ gg 🫡

TIL floating DC and ocean compute exists? Apparently it’s free cooling + optimal for places in tight geography… It’s going to be wild taking a cruise ship, and passing by a bunch of $NBIS or Samsung floating AI DCs one day. If this ever becomes a thing like orbital compute?

I forgot how good fable 5 is at creating landing pages. It understands webgl, scroll behaviors and text animations so well with simple prompts.