Sabitlenmiş Tweet
Everyone's debating which model codes better. Meanwhile OpenAI and Anthropic both dropped guides on the same thing: harness engineering. The model isn't the bottleneck. Your scaffolding is.
🧵
English
Alessio Masucci
348 posts
Alessio Masucci
@SucciMas
Survived jQuery, Angular, Vue, React. Now watching AI write my code. Principal Eng @immobiliare_it · Building @redirectsapp · Rome
Rome, Italy Katılım Ekim 2021
259 Takip Edilen51 Takipçiler
@trq212 Now the hard part: deciding what's worth a phone message and what needs your full attention
English
@tannerlinsley Type-safe routes, type-safe prompts. Makes sense
English
Claude 🤝 TanStack
Felix Rieseberg@felixrieseberg
A small ship I love: We made Claude.ai and our desktop apps meaningful faster this week. We moved our architecture from SSR to a static @vite_js & @tan_stack router setup that we can serve straight from workers at the edge. Time to first byte is down 65% at p75, prompts show up 50% sooner, navigation is snappier. We're not done (not even close!) but we care and we'll keep chipping away. Aiming to make Claude a little better every day.
English
@Cloudflare Inference where the workers live. One less hop, one less failure point
English
Kimi K2.5 is now on Workers AI, helping you power agents entirely on Cloudflare’s Developer Platform. Learn how we optimized our inference stack and reduced inference costs for internal agent use cases.
cfl.re/4bmpZgb
English
50 people. No press release. Just shipped
x.com/cursor_ai/stat…
Cursor@cursor_ai
Composer 2 is now available in Cursor.
English
@rdoar4 @mitsuhiko Fair point. And I say it as a daily Cursor user. The fact that you're thinking beyond the editor is exactly why I'm not going anywhere 😉
English
@SucciMas @mitsuhiko We are very proud of our editor, but our goal is to help builders ship incredible software, and the editor is only a part of that story (and needs to evolve drastically as AI evolves).
We have a lot of other things we're proud of that is not our editor 😉
English
Does anyone know if there is a way to use composer 2 via API on a non cursor harness?
English
Composer 2 vs Opus 4.6 vs GPT 5.4 - a totally unscientific test
> Create a Twitter clone. Use Better Auth, Vite, Sqlite, Drizzle, Typescript, and React with Tanstack Start.
Each one took ~5 mins in plan mode. Each had access to a browser to test.
Composer: 5 mins $6.04 1,250 LOC
Opus: 19 mins $10.43 1,000 LOC
GPT: 22 mins $14.15 2,000 LOC
Opus seems to use the cache WAY more than composer, so it's not really 10x more expensive.
Composer app ran first try. Other two needed a bit of CORS debugging but did work.
Code between all was extremely similar
All three done inside cursor - so Claude Code / Codex may have been different
Models used:
Composer 2.0 (regular, not fast)
Opus 4.6 Medium Thinking
GPT 5.4 Medium Thinking
Wes Bos@wesbos
Cursor just launched Composer 2 - their own model. It's 10× cheaper than Opus 4.6 and supposed to rival it. I've been using it for a few days, I don't have any skewed graphs to show you but from a pure vibes POV I can tell you it's pretty good™ My litmus test right now is if it can build a 3D Printable model with Manifold CAD. I build a Gif zoetrope generator and it did fantastic.
English
"Just use AI to write the tests." Sure. Now you have tests that pass and software that doesn't work. The test was never the point. Understanding the edge case was
English
Bold of you to pick a side in a debate that'll be irrelevant in 18 months
x.com/skirano/status…
Pietro Schirano@skirano
MCP was a mistake. Long live CLIs.
English
I didn't know you could disable Claude Code attribution when committing code.
To fix it, I asked Claude Code to disable attribution, and it updated the global settings. json file.
No more "Co-Authored-By: AI <ai@example.com>" comments.
Santiago@svpino
@Yuchenj_UW I really hate that Claude does this. I had to write my own skill + hook to prevent it from doing this.
English
@atmoio It's not designed, it's trained. RLHF on likes is a candy factory by definition
English
With LLMs we’ve basically invented digital candy.
LLM written content is designed to maximize engagement in a way human text will never be able to match. It’s why AI-written posts constantly go viral on here.
AI will win not because it’s better but because it’s more delicious.
English
Everyone is shipping "production-ready" now. Nobody is defining what production means anymore
x.com/GoogleAIStudio…
Google AI Studio@GoogleAIStudio
English
Senior engineers used to be valued for what they could build. Now they're valued for what they choose not to build. The skill set flipped and most people haven't noticed
English
@kentcdodds Cursor shipped a proprietary model that beats Opus at 1/5 the cost. The benchmark era is over, the "build your own" era just started
English
I've been using this model for a while and while it's not quite as good as GPT 5.4, it's much faster and cheaper
Cursor@cursor_ai
Composer 2 is now available in Cursor.
English
@emanueledpt Managing parallel agents from your phone via worktrees. The laptop is optional, the context window is not
English
Worktrees and Git Branches are available in the Codex Remote Control for iOS!
Remodex now has:
→ Worktrees
→ Handoff to Worktree
→ Worktrees in the sidebar
→ Create Branch
→ Switch Branch
→ PR flow improvements
→ Faster UI interaction
This was well requested as well!
More things coming! 👀🍴
Live in the new npm and app version!
Emanuele Di Pietro@emanueledpt
Worktrees are coming 👀
English
@cgtwts Still need someone to understand what working means. That part does not vibe-code itself
English
POV: vibecoders when they realise AI agents can write, debug and ship code for them 24/7
Google AI Studio@GoogleAIStudio
vibe coding in AI Studio just got a major upgrade 🚀 • multiplayer: build real-time games & tools • real services: connect live data • persistent builds: close the tab, it keeps working • pro UI: shadcn, Framer Motion & npm support we can't wait to see what you build!
English
Chose @convex over Postgres for redirections.app. No migrations, no ORM, no schema drift. The tradeoff is vendor lock-in. Six months later I'd make the same choice
English
@Google @GoogleAIStudio "Vibe coding" and "production ready" in the same sentence. Bold strategy
English
Introducing a new upgraded vibe coding experience in @GoogleAIStudio.
You can now turn any idea into functional, production ready apps. Build multiplayer games, collaborative tools, apps with secure log-ins and more.
English
"93% of jobs disrupted" sounds terrifying until you read the fine print. 10% of tasks automatable. 39% average exposure. Published by a $22B consulting firm that launched its "AI Factory" 48 hours earlier.
The disruption is real. The headline is marketing.
English
@AprilNEA Every AI company is three months away from accidentally becoming a cloud provider. Anthropic just got caught early.
English
🧵 I just reverse-engineered the binaries inside Claude Code's Firecracker MicroVM and found something wild:
Anthropic is building their own PaaS platform called "Antspace" (Ants + Space).
It's a full deployment pipeline — hidden in plain sight inside the environment-runner binary. Here's what I found 👇
English