Brandon Wie

A loving father and a loving husband. Life-long learner.
Full-stack AI Engineer @ |

[Liam Arzola 🦀 @liamarzola]

We finally have an OpenVMM based sandbox! MSR / AzSRG just open sourced nvx (https://t.co/X7ACDuMZOQ), a sandbox built to support most agent workloads while stripping out all unnecessary components resulting in the lightest hardware isolated uVM I've seen

[Chrome for Developers @ChromiumDev]

🔒 Prevent unauthorized data exfiltration with connection allowlists → https://t.co/CZNCNC0UiO

Third-party scripts or AI-generated code can bypass app-level checks to send sensitive information to unauthorized servers. Connection allowlists give you explicit control over network requests by making the browser the gatekeeper of all network connections originating from your page.

[Animate UI @animate_ui]

Animate UI 2.0 is out 🎉

Animated primitives for Radix UI, Base UI and React Aria, built with Motion.
Keep your shadcn/ui components.

✔ A CLI that animates them in one command
✔ An AI skill that animates anything (components, icons, interactions) and reviews your motion

[DeepSeek @deepseek_ai]

🚀 DeepSeek-V4.1-Flash-1017 is here!

Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.

🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.

1/4

[lauren @poteto]

your @bot is now just one button away!

[Brandon Wie @BrandonWie]

@bot @milichab @bot set this up

[Grok Bot @bot]

Grok Bot is now one step closer.

Set it as your iPhone's Action Button and press it to talk with Bot.

[MiniMax (official) @MiniMax_AI]

MiniMax M3.1 Flash: Space Performance + Bunny Speed

🚀🐰 Yes, Space Bunny - which has topped global leaderboards with 200T+ tokens - was an earlier version of M3.1 Flash.

💰Redefining the performance economics of agentic AI. M3.1 Flash achieves frontier-level intelligence across key tasks such as Cowork (Automation-Bench: 62%; ALE: 30%), Coding (DeepSWE: 74%) and Visual Reasoning (BabyVision: 91%), at 1/3 the cost of M3 and 3x the speed.

API pricing (flat rate): $0.15 / $0.60 per 1M input/output tokens, with cache reads at $0.015 / 1M tokens.

Model weights will arrive on Oct 23.

API: https://t.co/fHRdSV73Hr
M Plan: https://t.co/XfJWSyauda
MiniMax Code for desktop: https://t.co/SGmLTI2HQm

[Anton Martyniuk @AntonMartyniuk]

10 more books I can't recommend enough:

On your way to Software Architect (Part 2)

Here's the order I'd read them in again 👇

𝟭. 𝗙𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹𝘀 𝗼𝗳 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 (Mark Richards, Neal Ford)
↳ The map of architecture styles and the "-ilities" behind them.
↳ First law: everything in software architecture is a trade-off.

𝟮. 𝗣𝗮𝘁𝘁𝗲𝗿𝗻𝘀 𝗼𝗳 𝗘𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 (Martin Fowler)
↳ The reference for problems you hit in almost every app.
↳ After it, you'll spot these patterns in code you own.

𝟯. 𝗠𝗶𝗰𝗿𝗼𝘀𝗲𝗿𝘃𝗶𝗰𝗲𝘀 𝗗𝗲𝘀𝗶𝗴𝗻 𝗣𝗮𝘁𝘁𝗲𝗿𝗻𝘀 𝗶𝗻 .𝗡𝗘𝗧 (Trevoir Williams)
↳ Saga, CQRS, and API Gateway in real .NET code.
↳ The new edition is updated for .NET 10 and C# 14.

𝟰. 𝗗𝗼𝗺𝗮𝗶𝗻-𝗗𝗿𝗶𝘃𝗲𝗻 𝗥𝗲𝗳𝗮𝗰𝘁𝗼𝗿𝗶𝗻𝗴 (Alessandro Colla, Alberto Acerbis)
↳ How to apply DDD to the legacy monolith you already have.
↳ Step by step: from a big ball of mud to modules and microservices.

𝟱. 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗘𝘃𝗲𝗻𝘁-𝗗𝗿𝗶𝘃𝗲𝗻 𝗠𝗶𝗰𝗿𝗼𝘀𝗲𝗿𝘃𝗶𝗰𝗲𝘀 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 (Hugo Filipe Oliveira Rocha)
↳ The real pain of events: ordering, duplicates, eventual consistency.
↳ Shows how to keep data in sync when every service owns its own.

𝟲. 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲: 𝗧𝗵𝗲 𝗛𝗮𝗿𝗱 𝗣𝗮𝗿𝘁𝘀 (Neal Ford, Mark Richards, Pramod Sadalage, Zhamak Dehghani)
↳ How to split a system and its data when there's no best answer.
↳ Sagas, data ownership, and service size, with clear trade-off tables.

𝟳. 𝗦𝘆𝘀𝘁𝗲𝗺 𝗗𝗲𝘀𝗶𝗴𝗻 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄, 𝗩𝗼𝗹. 𝟭 (Alex Xu)
↳ A 4-step framework to design any system.
↳ Rate limiter, URL shortener, chat. Classic problems, clear diagrams.

𝟴. 𝗦𝘆𝘀𝘁𝗲𝗺 𝗗𝗲𝘀𝗶𝗴𝗻 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄, 𝗩𝗼𝗹. 𝟮 (Alex Xu, Sahn Lam)
↳ Harder cases: payment system, digital wallet, stock exchange.
↳ Worth reading even if you never plan to interview.

𝟵. 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝗶𝗻𝗴 𝗳𝗼𝗿 𝗦𝗰𝗮𝗹𝗲 (Lee Atchison)
↳ How to keep high availability as traffic and teams grow.
↳ Teaches you to measure and manage risk, not hope it away.

𝟭𝟬. 𝗦𝗶𝘁𝗲 𝗥𝗲𝗹𝗶𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 (Betsy Beyer, Chris Jones, Jennifer Petoff, Niall Richard Murphy)
↳ How Google keeps huge systems running: SLOs and error budgets.
↳ Architecture doesn't end at deploy. This book starts there.

📌 𝗢𝗻𝗲 𝘁𝗶𝗽 𝗳𝗿𝗼𝗺 𝗺𝘆 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲:
Before you read about a pattern, find where your project needs it.
Books stick when you have a real problem to solve.

Which of these have you read, and which one helped you the most?

[AI_Explorer @ai_explorer25]

15 AI Research Voices on X

@amaarora= AI papers explained through code.
@cwolferesearch= Clear, in-depth LLM research breakdowns.
@sedielem =Diffusion models explained deeply.
@MaartenGr= Complex AI made visual.
@seb_ruder =NLP and multilingual AI insights.
@Tim_Dettmers= Quantization and efficient LLMs demystified.
@arankomatsuzaki= Research papers distilled into takeaways.
@srush_nlp= Transformers explained with hands-on code.
@ai_explorer25= New AI tools, peoples and ai commentary
@rasbt= LLM research translated into practical code.
@natolambert= RLHF and post-training unpacked.
@eugeneyan= AI research meets real-world systems.
@sh_reya= AI evaluation beyond benchmark scores.
@HamelHusain= Practical, no-nonsense AI evaluation.
@ch402= Neural networks decoded from within

[0xMarioNawfal - The Roundtable Show @RoundtableSpace]

Cloudflare open-sourced the security audit skill that seeded its own vulnerability hunting harness. Your coding agent can now run it on your code.

How it works, in 6 phases:
→ maps your codebase
→ hunts for bugs area by area so nothing gets skipped
→ sends every finding to a fresh agent that didn't find it, to confirm it's real
→ writes structured results
→ verifies each record again, independently
→ writes the final report

You get a findings file split into confirmed, needs validation and rejected, so you fix the real problems first.

Install:
npx skills add https://t.co/ptOfJggqQB --skill security-audit

Then open your agent in your project and type: "security audit this codebase"

MIT licensed, 31.3K stars.

[schotek @schotek]

Odinstaloval jsem Docker Desktop a místo něj začal pro vývoj linux věcí z Macu používat nativní Apple Containers a je to supr - paměťová náročnost šla dolů, IO pro diskové operace je někde úplně jinde. Pro UI management doporučuju Davit...

[Hako @devswha]

Next release spoiler

[Brandon Wie @BrandonWie]

I tried @t3dotcodes for the first time after seeing a random post of someone talking about seamless subagent integration with other AI agent vendors, and it is actually including mobile app integration via t3 connect. Works smoothly with the personal harness as well. I'm gonna spend more time with it.

[Brandon Wie @BrandonWie]

the price seems like impossible to compete

[Cline @cline]

DeepSeek V4.1-Flash-1017 has released, beating GPT 5.6 Sol on Terminal-Bench while being ~17x cheaper.

It improves the previous DeepSeek V4.1-Flash result by 35%, accomplished with "no algorithmic innovation" but instead by RL training it on multiple coding agent harnesses.

[ClaudeDevs @ClaudeDevs]

Claude writes about 90% of Claude Code's code.

From June to September, merged PRs more than doubled while the share reverted fell by more than 40%.

We wrote about how we changed code review to get there, and how to adopt it:

https://t.co/9sZe64g3mb

[ClaudeDevs @ClaudeDevs]

We keep improving performance across Claude, and you may have noticed it starts responding sooner!

Claude Code now gzips the conversation it sends on every turn, so it reaches the model faster. Cloud sessions also start up as soon as you start typing.

[OpenCode @opencode]

GLM 5.3 Flash is now free in OpenCode

[Brandon Wie @BrandonWie]

the time is coming i guess..

[DeepSeek @deepseek_ai]

🚀 DeepSeek-V4.1-Flash-1017 is here!

Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.

🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.

1/4