Arjun | Building Stuff with AI

Next phase of Icarus: Personal Intelligence What if your computer didn't just optimize itself—but optimized you? It learns your routines, removes friction, and prepares everything you need before you ask. Less time managing your computer. More time creating.

Next phase of Icarus: Personal Intelligence Your computer learns how you work and prepares everything before you ask human optimization. Open your laptop, and your apps, creators, and workflows are already waiting for you. Still in early stages, Intent will become the future.

We're entering the next phase of Icarus: Personal AI What if your computer learned your habits and adapted automatically? If it knows you start your day with YouTube, Cursor, your favorite creators, and X accounts, everything is ready the moment you open your laptop. local-first

@0xCodez loop engineering is great one should checkout the last videos of anthropic they also discuss regarding this

A senior Anthropic engineer just dropped 11-page PDF on "Loop Engineering" for agentic systems. The shift: you stop prompting the agent. You build the system that prompts it instead. Schedule → Discover → Build → Verify → Repeat Every loop runs one turn, five moves: • Discovery: it finds its own work - failing CI, open issues, recent commits - instead of being handed a list. • Handoff: each task gets an isolated git worktree so parallel agents don't collide. • Verification: a second agent, told to assume the code is broken, reviews the first. The "thing that can say no." • Persistence: results get written to disk, never left in a context window that gets flushed. • Scheduling: an automation wakes it on a timer. That's what makes it a loop. The key insight: an agent grading its own work always praises it. This 11-page PDF changed how I'm building agentic systems today. Read it now, then explore the article below.

We believe the future of computing is intent, not apps. With Icarus, you can run isolated projects, install packages safely, and create automations for almost anything—all from a single interface. One screen. One AI. Follow me to see the journey of Icarus (still in production)

Loop Engineering is getting hype now. But not many talks about how to actually do it So I open-sourced the template my team uses to build agent loops: - a shared artifact / knowledge layer - logging, verification - and a codebase harness so work compounds across runs Plus a 20-min deep dive on how to think about it and set it up for real: github.com/AI-Builder-Clu… Copy the template. Adapt it to your own loops.

Today we have @openclaw that can control almost everything on a computer. But what if the OS itself was built around AI? No apps. No menus. No searching. Just tell your computer what you want, and it gets it done. I'm building Icarus. follow the journey of Icarus with me.

Hey do you guys know the future of OS? Would it be a completely new OS or a well Optimized AI Layer/app that would sit right there and you ask it what you want it would be done just simply in plain English or your native language. I am building it. Here in Public. Icarus

what are your thoughts on AI Operating layer / System ? A completely new OS for your system or a Layer that optimizes everything on your current OS. You just need to ask it, and it would be done, No other apps only intent. what would you do if you had following options:

Been quietly building something I'm really excited about. Forget searching through folders. Forget wondering which app to use. Just ask Icarus. Icarus is an AI operating layer where you tell your computer what you want, and it gets it done. follow me and see journey of Icarus.

Been quietly building Icarus. An AI operating layer where you tell your computer what you want, and it gets it done. System and memory management are done. Still a long way to go, but we're making progress. Drop a like and Follow along for updates.

After a long break, I'm back to building I'm building Icarus — an AI operating layer where you tell your computer what you want and it gets it done. Phase 1: Optimized File system, scheduler, agent runtime, AI shell, Memory and context-intent aware. This is the future of OS

After a long break, I'm back. I'm building Icarus — an AI operating layer where you tell your computer what you want and it gets it done. Phase 1: Optimized File system, scheduler, agent runtime, AI shell, Memory and context-intent aware. The future of OS is intent

Back in builder mode. I'm building Icarus - an AI operating layer for the computer. Just tell your machine what you want, and it plans, reasons, and executes. Phase 1: Optimizes file search, schedulers, memory, AI Shell. Stay tuned for future of OS that works on intent.

@techNmak this is good but you can achieve same result with a well documented issue , well written prompt and a correct md file to help your agent around

A dev got so frustrated watching his AI agent write 500 lines for a 5-line problem that he built a fix. He called it Ponytail. Named after the guy every team has - long ponytail, oval glasses, been there longer than the version control. You show him fifty lines; he looks at them, says nothing, and replaces them with one. Now your agent does the same. Before writing anything, it looks for a reason not to. 80-94% less code. 47-77% cheaper. 3-6x faster. The best code is the code you never wrote. GitHub Repo: github.com/DietrichGebert…

@martinboksa this is so awesome refreshed some childhood memories

I made this multiplayer bomb arena in 1 day with /goal + Claude Fable 5. Crazy. You can play it now at blastmates.app Real-time multiplayer. Drop bombs, grab powerups, blow up your friends. Rooms for 4/6/8 players, bot opponents, mobile support (a bit rough but it works 🤷‍♂️), and a sudden-death mode that spirals the walls inward if a round drags on. - Coded entirely in @claudeai with Fable 5 - All graphics generated with Fable 5, every sprite, the whole UI - Music generated with Fable 5 too - The only thing I touched by hand: a few .env files Built solo, in a day. Blow up your buddies 💣

🌘 Meet Kimi K2.7 Code HighSpeed! A high-speed mode of our latest open-source multimodal coding model, Kimi K2.7 Code. ⚡️ Up to 6× faster: Around 180 tok/s on coding tasks with median-length inputs, and up to 260 tok/s on shorter-context tasks. 🔷 Rolling out to Kimi Code Beta Program members, Kimi API developers, and Kimi Business users. (Access will remain limited for now due to capacity constraints.) 🔷 No invite needed. Anyone who joins the Beta Program has a chance to get access 👉 kimi.com/code/beta Open intelligence should be instant, affordable, and borderless. We'll continue improving the model and expanding access as more capacity becomes available! 🔗 Kimi Code: kimi.com/code 🔗 API: platform.kimi.ai

Try it out! (Partially trained on Colossus 2)

@anujcodes_21 Yo this is awesome

🚨Anthropic just showed a 24-minute workshop on how to actually do prompts for Claude. Taught by the people who built it. Free. No registration. No paywall. I've seen $300 courses that don't cover what they teach in the first 8 minutes. Watch it and bookmark it now!