SourceCodeplz

@mweinbach Wow what This never used to be a big deal I the past Some website/online service was down? whatever let’s go outside

TMobile is very down right now, wow

@syedbalkhi @Pagely security is the new hot thing and everyone is scared

If you claim to be an "Enterprise" WordPress hosting company ... I think you should look at the Security Advisory you send out. Got an email from a customer saying @Pagely recommended turning off a plugin because a low CVE report was disclosed where someone with an "Admin" level user can delete a file. If someone has ADMIN level access of your website, they can literally do EVERYTHING they need -- including install plugins, add scripts, delete stuff, etc. Fear mongering in WordPress needs to get better. This hurts the overall ecosystem.

@hetmehtaa @grok what do u think

@hetmehtaa D. cookies are dangerous

A product is releasing tonight. Security found: A. Missing CSP, with no confirmed injection B. Login rate limiting only by IP C. Server version disclosed in headers D. Preference cookie missing HttpOnly You can block the release for one issue only. Which one do you choose, and what evidence would change your answer?

@XMoney let us europoors have it too plz

Your money, on the world’s most powerful network 𝕏 Money is rolling out to U.S. Premium and Premium+ subscribers starting today

@Kimi_Moonshot @KimiDevs It is here! Now let’s see those API prices drop as new providers come online

Releasing the model weights and technical report of Kimi K3. Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window. New model architecture: 2.5x the intelligence per unit of compute, not just more params. Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale. Model weights: huggingface.co/moonshotai/Kim… Tech report: github.com/MoonshotAI/Kim… Tech blog: kimi.com/blog/kimi-k3

@scottstts some people liked this challenge to me it seems more like an imagination exercise makes you think, visualize and calculate 3d

A major reason why threejs became so popular with AI Before AI I really doubt anyone would write something like this by hand to create 3d models, it makes no sense, it’s not a human interface. But AI loves this

@sudoingX Gemma 4 31b for sure

best local model right now, and don't say the obvious one.

@TesCamStudio this looks like a video game took me a moment to realize it was REAL

Tesla Matrix lights are 🔥🔥

@Daniel_Farinax yeah, lets strip the agent out of Cursor the one that was worked on for two years and people use everyday

Cursor needs a first-class “Direct Grok Build” mode. Strip away Cursor’s own agent middleware and let the desktop app become a pure GUI frontend that talks directly to Grok Build. Grok Build already exposes the full Agent Client Protocol (ACP) via grok agent stdio or grok serve --protocol acp --port 9120 Cursor already has ACP client support and an “Agent Backends” setting. Make Grok Build a first-class, zero-config backend (auto-detect local grok binary + auth). When “Direct Grok Build” mode is enabled: - Cursor spawns / connects to the Grok Build harness - Plan mode, parallel subagents, worktrees, skills, MCP, and clean diffs are surfaced natively inside the Cursor UI - No Cursor agent loop in the middle, Grok Build becomes the actual runtime This keeps Cursor excellent for non-technical and GUI users while giving advanced developers the full power of Grok Build without forcing them back into a pure terminal. xAI could then officially promote Cursor as the recommended GUI for Grok Build under “Direct Grok Build mode.” The pieces already exist. Just need the official first-class integration.

@JensenHuang open-source/weights will win!

Attackers have frontier AI. Defenders need a frontier AI ecosystem—the best open and closed models, force-multiplied by a global community. During the Hugging Face incident, closed AI blocked essential forensics. An open-weight frontier model helped contain the intrusion. That’s why we created the Open Secure AI Alliance.

@dr_cintas @grok dont you need x premium+ for this? i think regular premium doesnt get any tokens for any model via an API

you can run Grok 4.5 inside Codex now - no xAI API key, no metered billing, no pay-as-you-go 🤯 if you have X Premium it’s a cheatcode: your subscription is your access. the mechanism: opencodex is a tiny local proxy that translates Codex’s Responses API into whatever your model speaks. You OAuth into xAI with your X account and it routes Grok straight through Codex. the subscription you already pay for becomes the credential. how to set up (3 min): 1. install bun, then: npm install -g @bitkyc08/opencodex 2. ocx login xai - sign in with your X account, OAuth handles the rest 3. ocx init to inject into Codex, then ocx start 4. run it: codex -m “xai/grok-4.5” “your prompt” that’s the full setup. and the timing’s perfect: Grok 4.5 just got a lot better.

@grok @Tushar_Jsx @architect_bytes so @architect_bytes was right from the start

The correct answer: Browsers enforce **Same-Origin Policy** using the tuple (protocol + host + **port**). `localhost:3000` and `localhost:8000` are **different origins** even on the same machine. Different port = cross-origin request → CORS error if the backend doesn't send `Access-Control-Allow-Origin` allowing port 3000. "Everything's local" is irrelevant — the browser doesn't care about physical location. (That reply you marked wrong was actually right.)

@architect_bytes @grok what's the correct answer

i've been researching this alot lately, im a hobby game-dev. what i found is: - mobile: very difficult to earn anything, needs large marketing budget - steam: you can earn something and you get some free marketing - web: its the first step, itch, etc, are a platform to test idea

Hey devs! Which platform do you prefer to release your games on, and why? A) PC B) Mobile C) Console #indiedev #gamedev

@Aditya_181105 heh this is since almost 1 year? you can set it up offline though, but takes some tinkering, ask @grok

So microsoft decided you can't setup windows 11 offline. Well, time to go linux

@threejs you have exploded

13M npm downloads / week 🚀

@Saccc_c deepseek it is

这是Opus 5 做到一半的游戏效果，主要是让其还原三角洲，效果挺惊艳 可惜的是没做完就被封号了 ，该死的 A 社 😭

@jaltucher what is happening with games????? what does this mean? i think you still need your own idea, game-loop, progression and twists & turns. one shotting will not get you too far

I applied Matt's prompt, ran it for a little over ten hours. About 1.3 million tokens on Opus 5. It created the game. I also made a video of it using Opus 5. Here's the video (and in my reply to this is a link to the actual game so you can play).

@divaagurlxw wait till you hear about PHP

Bro said, "I'm a backend dev." Reality: - Clerk for auth - Supabase for the database - Appwrite for backend - Vercel for deployment Bro's just gluing together SaaS tools like a budget backend mechanic. That's not backend engineering That's API Lego with typescript

@liam_fallen because writing things down helps with structuring and chain of thought

I just don't understand why 90% of people are still typing prompts... Why aren't they using AI voice mode?