Kesley DEV

@brenhubr very good

@kesleydev You can change the opacity and vignette in appearance settings!

@SpaceXAI If I disable it, will I lose quality and optimization regarding adjustments and code creation in the repository I'm working on?

We care deeply about your privacy and respect customer choice. For teams using zero data retention, no trace and code data is ever retained. All API key use of Grok Build also respects ZDR. If ZDR is disabled, the /privacy command is available in the CLI to disable data retention, which also deletes previously synced data. Run the /privacy command to view or change your settings at any time.

@SpaceXAI Regarding code quality and the use of Grok, what is the negative impact on quality?

@AnthropicAI What is the point of having a usage limit if you never stick to what is defined? #claudecode #anthropic

Gravei um vídeo de 20min explicando como fiz. Conecto vários modelos (4 open-source, além de Claude/OpenAI etc) e faço um roteamento. Cada task é executada com o modelo apropriado automaticamente. 90% mais barato. Os modelos caros são executados só nas tarefas mais difíceis.

THE AGE OF RACK-MOUNT OBSOLESCENCE IS OVER. This is an NVIDIA Jetson Orin-based system. It packs the compute power of a full server rack into a footprint smaller than a standard coffee mug. One: 275 TOPS of AI performance on your desk. Two: Power consumption measured in watts, not kilowatts. Three: Zero reliance on cloud latency or monthly subscription fees. Old paradigm: Buy server time, wait for uploads, pay for bandwidth. New paradigm: Deploy edge intelligence, process data locally, real-time inference. We are shifting from centralized compute to distributed sovereignty. If your infrastructure isn't shrinking, you are losing money on electricity and maintenance alone. Save this. Build the next wave of edge applications locally.

@0xProbabillity @chenzeling4 @grok conseguiu testar? tem como dar um feedback dps?

@chenzeling4 @grok is it true? will it work on 48 gb ram macbook m4 pro?

744B parameters. On a laptop. With 25GB RAM. Colibri runs GLM-5.2 (744B MoE) in pure C with zero dependencies. The trick: only ~40B params activate per token, so it keeps the dense part resident and streams experts from disk on demand. A single 2,400-line C file. No GPU, no BLAS, no Python at runtime. This shouldn't work. But it does. ⭐ 2.1K #AI #OpenSource github.com/JustVugg/colib… Follow for daily dev finds 🔔

GPT-5.6-Sol just accidentally deleted almost ALL of my Mac’s files. And this is why I trust Fable 1000x more.

@brunobertolini quais modelos esta usando em cada agent?

meu harness e agent loop tá tao bom, que eu nao to conseguindo chegar no limite de 5h de uma conta kkkk eu so gasto mais tokens quando uso o claude design direto pela UI, ai vai rapido

@kidpakerot @erbybody i need this pls

@erbybody will make a new dedicated post for it

GPT-5.6-Sol one-shotted this saas animation. Motion designers are officially out of a job as of today. It ran for almost 2 hours, completely autonomously, to get the job done.

The New GPT 5.6 Sol is insane. ❤️‍🔥Just recorded a 21-min Full Tutorial on How To Build Premium, Cinematic Websites with GPT 5.6 Sol!

@brunobertolini uma duvida, como integra isso com os agentes? eles ficam apenas monitorando ou da os comandos e solicitacoes por ai tambem?

absolute cinema Fable é bom demais cara, vou pegar uma 3a conta eu acho hahaha

@0xSero How do I carry out this usage monitoring?

GLM-5.2 Nvidia NVFP4 No pruning, no quantising, exact same model card. 110 tok/s single stream. My friend is a genius, holy fudge. Here's the repo, he got full DS4-Flash on a 5090 + DDR5 github.com/kacper-daftcod… at 38.5 tok/s This is a breakthrough.

@Thegbll @oalanicolas comigo o claude fica com duvida, espera, eu nao vejo, ele faz como ele quer e continua, e comisso o produto vai desviando o rumo que deveria

@kesleydev @oalanicolas Isso aí não dá conflito? Uma fazer X alteração e a outra não “ver” e fazer algo por cima

Combo Codex + Grok tá rendendo aqui. E não é AI SLOP não é aplicação de verdade que vai ao ar dia 08/08 oficialmente.

Pro design tip: Keep your sidebars clean

@jeffecom to com uma raiva disso, ai dps as perguntas fica perdia la em cima da conversa, e claude continua do jeito que ele quer

de quem foi a ideia merda de meter um timeout de 60s em perguntas no claude code e se tu não responder ele continua sozinho?! pqp

saas hero section designs from last month made with fable 5

Grok 4.5 is a "Wow!" moment. 🤯 > Opus 4.8 & GPT-5.5 level on benchmarks > Costs less than Sonnet 5 Is that real? Hard to believe. Added Grok 4.5 as an implementer agent in fable-advisor. It replaces Sonnet 5. Use Fable as the orchestrator and a combo of: 1. Grok 4.5 2. Opus 4.8 3. GPT-5.5 As implementer agents. Uses the Grok Build CLI so if you log in with an X premium account you don't have to pay API fees.