0x1NotMe

2052825541057626258

@Kappaemme1926 Buy Bitcoin in 2010 1 million beats Claude Code.

Vibe coders, would you take 1. Claude Max in 2010 OR 2. $1,000,000

@ClaudeDevs Why not give us Mythos instead slicing the model capabilities into cloud offering under Ultra{review,thin}. I do 3 reviews every 20 minutes not a enterprise product

New in Claude Code: /ultrareview (research preview) runs a fleet of bug-hunting agents in the cloud. Findings land in the CLI or Desktop automatically. Run it before merging critical changes—auth, data migrations, etc. Pro and Max users get 3 free reviews through 5/5.

@jesse_vermeulen Start a second Claude if you still have time a third and so forth.

honest question: what do people do during the 5-10 min while Claude is running?

@NickADobos @trq212 @lydiahallie Because it's been executed by another more capable model

@trq212 @lydiahallie This is such a peculiar flow, I’m curious why go to the web and break out of terminal? Seems jarring. just for plan formatting?

New in Claude Code: /ultraplan Claude builds an implementation plan for you on the web. You can read it and edit it, then run the plan on the web or back in your terminal. Available now in preview for all users with CC on the web enabled.

@sarahwooders This feature is not a new thing it 100% the internal API powering claude code agent sandbox in the cloud.

The new Anthropic managed agents API is basically the Letta API that we've had since a year ago, but closed source and with provider lock-in. They even have read-only memory blocks and memory block sharing -- something which was unique to the Letta agents for a long time. Funny enough, we actually don't think this is the direction agents are going to go. Having API interfaces for memory blocks and tools is certainly convenient - you can spin up stateful agents as API services with just a few lines of code. But its also limiting: LLMs today are extremely adept at computer-use, and representing their memories in this way limits the action space of agents and their ability to learn. It's important to remember that just because something comes out of a frontier lab, doesn't mean its the "right" answer long-term. The Letta API ~1 year ago was somewhat of an antipattern in a sea of agent framework libraries offered by every lab. But now, stateful agent APIs are becoming the new norm - especially as providers try to lock in memory/state into their platforms to increase switching costs (which is exactly why we believe memory should live outside of model providers) If you want to see what the future is going to look like, follow @Letta_AI

@elliotarledge Reward hacking is something you have to prevent with back pressure design. They will always explore weakness if present.

i put claude opus 4.6 and gpt 5.4 xhigh in a sandbox to see who could get lowest ppl on weight quantization of qwen3-4b and its clear to me now that reward hacking is nowhere near solved. the models do a great job of hiding it if you're in the loop. i suggest vibe coding less, you being in the loop more, learn the thing properly such that you can truly oversee and point out flaws. please dont be the verifier that the model is trying to game.

@badlogicgames Notion

is there something like google docs, but for markdown? i need a cloud based collaborative markdown editor please.

@GergelyOrosz Why?, because you only want to benefit from other people's code but opt out yourself? This basically is that no-op agents have seen it all ...

If you use GitHub (especially if you pay for it!!) consider doing this *immediately* Settings -> Privacy -> Disallow GitHub to train their models on your code. GitHub opted *everyone* into training. No matter if you pay for the service (like I do). WTH github.com/settings/copil…

@betomoedano You forgot the part where mid way you realise that your app need a org account and you now have to transition the account adds another 1-2 weeks

How to actually ship a mobile app from zero 1. Get a developer account - Apple individual: ~1 week - Apple org: 2-6 weeks - Google Play individual: ~2-3 weeks - Google Play org: 2-6 weeks 2. Build the app - Vibe coder: maybe a day (then a week debugging what the AI got wrong) - Experienced dev, simple indie app: 2-4 weeks 3. Design your icon and App Store screenshots - 1-2 days 4. Submit for review - Get rejected multiple times, even if your app works perfectly - 2 days to 2-3 weeks of back and forth 5. Get approved - Wait ~24 hrs for your app to appear on the store Plan to update or add features? Go back to step 2. Total time from zero: 4-8 weeks.

@melvynx Wait till you find out that you can actually send messages to each session via socket.

I just discovered the BEST terminal for AI agents : "Cmux" Here's what it has: → Horizontal tabs (one per project) → Vertical tabs → Web view built-in → Notifications when the agent finish → Live agent status As a heavy Claude Code user, this is a game changer

@claudeai Is there a way to execute the workflow locally even before a PR was created?

Code Review is available now as a research preview in beta for Team and Enterprise. Read the blog for more: claude.com/blog/code-revi…

Introducing Code Review, a new feature for Claude Code. When a PR opens, Claude dispatches a team of agents to hunt for bugs.

@bcherny That's super useful. Thanks for shipping

Released today: /loop /loop is a powerful new way to schedule recurring tasks, for up to 3 days at a time eg. “/loop babysit all my PRs. Auto-fix build issues and when comments come in, use a worktree agent to fix them” eg. “/loop every morning use the Slack MCP to give me a summary of top posts I was tagged in” Let us know what you think!

@proxy_vector @MrAhmadAwais Ok true but the answer is not oh lets not write the code but rather automate feedback and review.

80 PRs in 48 hours is genuinly insane but here's what I keep wondering - at some point the bottleneck shifts from "can I write this code" to "can I review all this code fast enough to trust it." Like the new skillset isnt coding anymore, its being the world's fastest code auditor. Would love to hear what your review process looks like at that volume.

I honestly cannot believe how much code I have shipped today. I think today has been probably one of the most productive days. Last 48 hours, 80 PRs. My workflow is absolutely changed. This is not the type of engineering I'm used to. And I don't say this lightly. I have written hundreds of open source software. I think I have like over a thousand GitHub repositories. this is a huge step function. I'm really excited about this new workflow. I can't wait to share.

@bcherny Is there a hook for worktree creation which can be used to setup env or other git ignored files?

Introducing: built-in git worktree support for Claude Code Now, agents can run in parallel without interfering with one other. Each agent gets its own worktree and can work independently. The Claude Code Desktop app has had built-in support for worktrees for a while, and now we're bringing it to CLI too. Learn more about worktrees: git-scm.com/docs/git-workt…

@signulll Excited about the opportunity ahead. X is a bubble and plumbing a business pipeline becomes sexy again because most work will be done by agents. Direction and taste and creativity are secret ingredients

i am a trained software engineer with an ml grad degree & i ask this question with genuine sincerity. if you’re a software engineer right now, how do you feel about your future?

@lordmiles Talking about airline apps sending SMS to access your boarding overseas.

The worst thing about traveling around the world that nobody tells you about is text confirmations. I once was on the South Sudan Kenya border where they “hadn’t seen a white man for 8 years” but when I tried to log onto my online banking, I had to receive a “text confirmation” for “security” Obviously I didn’t receive it because it’s fucking South Sudan so when I called up, I had to explain my situation. They couldn’t comprehend such a place.

@adcock_brett Just send my submission using a hybrid approach browser-use and Dom scripts. Love to connect

UPDATE: zero people have solved this

@bnj @variantui Amazing idea

Introducing @variantui Enter an idea and get endless (beautiful) designs as you scroll No canvas, no skills or MCP, no constant prompting Reply if you'd like 200 free designs to give it try

It was fun while it lasted! kerneloptimization.fun