Your curated collection of saved posts and media
Introducing Code Review, a new feature for Claude Code. When a PR opens, Claude dispatches a team of agents to hunt for bugs. https://t.co/AL2J4efxPw
Rosy-faced Lovebirds on Saguaro cactus 🌵 https://t.co/hIrjuaFbfp
Rosy-faced Lovebirds on Saguaro cactus 🌵 https://t.co/hIrjuaFbfp
A shepherd in abandoned Sacred Grove, Italy, ca. 1952 - by Herbert List (1903 – 1975), German https://t.co/O6DXE1EXAg
A shepherd in abandoned Sacred Grove, Italy, ca. 1952 - by Herbert List (1903 – 1975), German https://t.co/O6DXE1EXAg
scrolling through this site feels like watching a movie → https://t.co/a41OXXNDFB https://t.co/jvN7Z1LVvM
"If Trump can depose any leader he likes, what is to stop Russia or China acting similarly against their foes? (Russia, of course, already has tried, in its war on Ukraine.) Is that the world we want to live in?" https://t.co/lvjGKwuJmv
today we're launching struct - the on-call agent every team deserves. and we're making it free. no sales call. 5 min setup. 80% faster on-call. https://t.co/vuX7iATL2l
Super excited to present MetaClaw 🦞 — our self-evolving agent that learns just by talking to it. 💬 Conversations → training trajectories 🧠 Failures → new reusable skills ⚡ Weights update live (no interruption) ☁️ Powered 100% via API — no GPU cluster needed Try it here 👇 https://t.co/p5yqdWmwzR Huge kudos to the team @richardxp888, Jianwen Chen, @Xinyu2ML, @lillianwei423, @StephenQS0710, Zeyu Zheng, @HuaxiuYaoML
@Yuchenj_UW Codex is a know issue :( It basically don't work with autoresearch sadly, in the way it's set up atm: https://t.co/4xjdf4wQyX I pung a friend at OpenAI to see if something can be done, e.g. need a /loop equivalent or something like that. More generally, I really dislike the -p + ralph loop pattern of running agents "headless". I want nice, interactive sessions running in tmux so that I can see what they are doing, pitch in, etc.
Andrew Ng just solved one of the biggest problem with Agents. He released Context Hub, a CLI tool to fetch live API documentation. One command. The agent gets exactly what it needs before writing a single line of code. Agents trained months ago are flying blind. They invent parameter names. They call functions that no longer exist. They confidently write code against a spec that changed in the last release. > No more hallucinated parameters > Docs pulled fresh before each call > Agents log useful discoveries > Notes persist between sessions The agent runs a CLI command before touching the code. Instead of relying on stale data, it reads the actual spec. Fast-moving APIs used to mean maintaining a doc dump in every prompt. Now the agent does that work itself. When it finds a workaround, it saves a note for next time.
https://t.co/e7xglk9wXC
🚨 New: Integrating Harbor (@harborframework) for end-to-end Computer-Use evaluation(for Windows and Linux) at scale with @thinkymachines' Tinker, OSWorld, @daytonaio, and bare-metal servers. We just added support for Computer Use, @tinkerapi, and OSWorld to Harbor - a framework for evaluating agents and generating RL training data by running large-scale rollouts across parallel sandboxed environments and collecting trajectories for SFT and RL. Repo and blogpost below 👇
New Tool for Immersive Filmmakers, Spatial Video Creators, and XR Developers: I’ve spent years teaching immersive creators how to make 360, VR180 and spatial video actually work across platforms. One problem keeps coming up again and again: the final delivery workflow is way more painful than it should be. So I built something to help. Today I’m sharing Hugh Hou Spatial Injector, a native macOS app designed to make it easier to prepare immersive video for YouTube VR and Apple Vision Pro in one simple workflow. What it does: ✅ Inject Google Spatial Metadata for YouTube VR without recompressing your video, so your original exported quality stays intact ✅ Convert your footage into Apple Spatial Video (MV-HEVC) using Apple’s native hardware encoding pipeline ✅ Simple drag-and-drop workflow with no Terminal and no outdated patchwork tools ✅ Helps creators deliver the same immersive story to more viewers across platforms This tool is 100% free. No Ad, no in-app purchase BS, just free and open-source. Example VR180 processed with the tool: https://t.co/X6mSBY5NmM 👇 Free Download link: https://t.co/D2M6PHHOTp I built this out of love for the immersive community and the belief that better tools help more creators bring powerful stories into 3D Spatial. If you know filmmakers or developers working in 360, VR180, Immersive 180, or Vision Pro, please share this post with them so more creators can benefit from the tool. Let’s keep pushing immersive storytelling forward together. #HughHou #AppleVisionPro #GoogleSpatialMetadataInjector #ImmersiveVideo #VR180
Surprises: - Perplexity higher than you'd think - removebg the 16th most popular AI tool - Google AI studio beating Lovable, and together they're ahead of all other vibe-coding platforms - JanitorAI is not what you expect - Three in the top 20 are AI chatbot friends - Google has four products in top 50
Meet GitClaw - the multi-model git-native @openclaw alternative. We set out to build an OpenClaw equivalent for enterprises (codenamed @lyzr__ai Chimera). One that is safe, with built-in version control, and that runs under strict controls instead of spontaneously launching tools, agents, and skills without oversight. Today, we hit an important milestone. GitClaw is the OpenClaw alternative built on the open agent standard GitAgent. That means versioning and file system management follow the same proven model software teams have trusted for decades. The core idea of Git, applied to agents. Now add vision and speech on top of that and you get something very powerful. GitClaw is multimodal. It can handle text, images, and even video. You now have a system that is more capable than OpenClaw and infinitely more controllable for enterprise deployment. Run it completely locally on your own machine or on a secure server. And very soon, GitClaw will be available inside Lyzr Agent Studio as an agent framework. Check out the video below to see our very own @Shreyaskapale interacting with GitClaw in real time. This is what enterprise-grade agentic AI actually looks like. GitClaw and GitAgent are both open-source. GitClaw: https://t.co/ORuMetTZB6 GitAgent: https://t.co/wUvEGNNivQ
We could tell you about how Copilot CLI can optimize your time, but we'd rather have @cassidoo sing about it instead. https://t.co/TJznR66Hof
Introducing the new App & Agent Rankings ✨ A better way to explore the AI ecosystem: - Apps can now rank within categories, or make trending - Request-count rankings, in addition to tokens - Apps can easily opt in and categorize their own generations. https://t.co/spwXlpfjZm https://t.co/oVqdiGUc4r
Today, we’re excited to launch Proactive Agents, a new standard for the AI concierge. This is a major step forward in helping businesses move beyond reactive support to building customer relationships that grow stronger over time. Here’s what’s new: 🧠 User Memory: Deliver personalized experiences to every customer. 📞 Outbound Voice: Build stronger relationships with proactive engagement. 🛠️ Agent Workbench: Debug agent workflows faster with an AI assistant. Throughout this week, we’ll be diving deeper into each of these features and how our customers are using them to deliver concierge experiences. Learn more about the full launch in the thread.
I’m excited to announce that @humanicai has been acquired! We built Humanic to capture how the best sales teams think, decide, and act to translate that into actionable context. Enterprise sales conversations deserved a true decision layer for automated judgment, and after 2M+ minutes of deal conversations, we’re proud to have set that in motion. Growing up in India, @abhi_is_agi and I admired the behemoths of Silicon Valley, but joining forces with one was not on our cards. It’s an honor. I wish I could divulge more, but our lawyers say we have to wait. We want to thank our earliest believers, our users, and the ones who stayed. It's still day one here in SF, and we are just getting started! See you in the next chapter.
Video just became native to AI agents. Today we’re launching HeyGen MCP. Now you can create HeyGen videos directly inside tools like Claude Web, Claude Code, Gemini CLI, and Cursor. Just connect your HeyGen account via OAuth and generate videos using your existing plan. We’ve put together a quick walkthrough showing how to set it up in Claude 👇
Grok is officially the #3 most visited Gen AI site in the world surpassing both DeepSeek and Claude The progress xAI has made in just one year is insane - from literally nothing to #3 worldwide Grok: ~314 million visits (up from ~271 million in December 2025 - fourth straight month of growth)
These attacks, like all phishing, rely on social engineering. Attackers impersonate trusted contacts or services (such as the non-existent “Signal Support Bot”) to trick victims into handing over their login credentials or other information. To help prevent this, remember that your Signal SMS verification code is only ever needed when you are first signing up for the Signal app. 2/4
Grok Imagine is the only one that maintains the style throughout the entire extension. I used to only be able to keep 50% of the extension with others solutions, but with Grok I have no breaks and perfect fluidity. The Baroque style represented by Grok Imagine. I think Grok's doing a great job.