John Eliel Baladjay

I run operations at a small semiconductor supplier, mostly through a personal AI system I built. Daily notes on what works and what breaks.

[John Eliel Baladjay @JohnECBal]

Every time I correct my AI, it writes one line the same turn: what went wrong, why, and where the fix belongs.

A correction that stays in the chat disappears with the chat.

[Wilton Wang @Wilton_6996]

永远不要指出你身边人的任何问题，包括你的亲戚和朋友。看透而不说透，看见而不拆穿，知人而不轻易评判。

一个真正成熟的人，渐渐会明白：不是所有的问题都需要指出，不是所有的错误都需要纠正，更不是所有的道理都值得当面讲出来。

亲戚有亲戚的难处，朋友有朋友的局限。每个人都有自己的认知、习惯和人生轨迹。你眼中的缺点，未必是别人眼中的问题；你认为正确的道理，也未必是别人当下能够接受的答案。

有些话，说出来是直率，听起来却是伤害；有些提醒，自以为是善意，实际上却是在冒犯别人的自尊。

成年人最高级的清醒，不是看不见别人的问题，而是看见了，依然懂得尊重；不是没有表达的能力，而是拥有克制表达的修养。

当然，不指出问题不等于纵容伤害，也不意味着放弃原则。涉及底线、责任和安全，该提醒时仍要提醒，该拒绝时必须拒绝。

真正的智慧是：
看破不说破，是修养；知人不评人，是格局；守住边界，是智慧；管好自己，是修行。

人这一生，少一点纠正别人的冲动，多一点反观自己的觉察，关系才能长久，内心才能安宁。

[mengtao wang @aiZttt__]

🚨 Grok 订阅福利：有机会免费领取 1 个月！

X / Grok 目前出现了新的订阅优惠：

Premium：首月 $0
Super：首月 $50（原价 $100）
Ultra：首月 $100（原价 $200）

邀请链接：
https://t.co/EKzm6Xjksi

使用 Google 账号、美国或印度地区网络环境，有机会领取。

⚠️ 优惠因账号和地区而异，不保证成功。务必确认结账金额为 $0，注意次月自动续费。

[Niko @Jcryptdefi]

thank me later

free 1 month of x premium

if you haven't claimed yours yet, now's the time

expires november 30, 2026

redeem at: https://t.co/ewdPpVSi43

[Mehmet @avuntusaati]

Dünyada hiçbir şey sandığınız kadar stres yapmaya değmez. Çoğu kaygı zamanla anlamını yitirir, çözümler kendiliğinden ortaya çıkar. Hayat her şeye rağmen akışına devam eder. Siz içinizi ferah tutun, yeter.

[知白 @HMingcoin]

心理学上有个词叫：螃蟹效应。

永远要记住，和周围人搞好关系的秘诀就是，不分享喜悦、不炫耀成功、不说三道四、不假装聪明。

[Google Antigravity @antigravity]

Browser use has been rolling out in Antigravity, closing the loop between researching, writing code, and verifying it live.

You choose what to log into, downloads pause until you approve them, and the agent can use the browser to run multiple tasks at once. Give it a try today.

[Valor Mental 🧠 @valormental]

Uno de los mayores trucos de la vida es no ofenderte nunca. Entrénate para tener la piel dura. No te lo tomes todo como algo personal. Deja que los demás no estén de acuerdo contigo. Si te ofendes fácil, te manipulan fácil. Si valoras tu paz mental, no te ofendas.

[John Eliel Baladjay @JohnECBal]

I use Claude Code, Codex and Antigravity for the same work. My rules live in one plain text file that all of them read.

My saved procedures sit in one folder, and each app links to it. Switching apps costs me almost nothing, because the system doesn't live inside any one app.

[Qoder @qoder_ai_ide]

Qoduck has earned its name. Today it upgrades to Qwen4-Max-Preview.

For 4 anonymous days it was a Qwen test build for Qoder users, updated daily.

Now live first on Qoder: every product, every plan, Teams and Enterprise included.

Limited time: 80% off, 0.1x credits.

[Qwen @Alibaba_Qwen]

Qwen4 is coming.

Deeper knowledge. Richer world understanding. Stronger coding and agentic capabilities.

Built for real-world tasks — from complex coding and knowledge work to front-end builds and professional software workflows.

Qwen4-Max is now in preview on Alibaba’s Token Plan. You can also get early access to the new model on QwenWork, Qoder, Qwen APP, and QwenStudio.

Put it to the test on a real problem. Show us what you build. 👀

Token Plan: https://t.co/LOB2Jdt3Sk

[OpenAI @OpenAI]

Now in beta: Businesses in ChatGPT.

Claim your business, manage your information, and get insights on how people engage with you.

Eligible businesses can also build a Business Agent to answer customers’ questions and help them explore what you offer.

https://t.co/N2GEZX0mFT

[Nan Yu @thenanyu]

PSA since it took me a second to discover it:
If you wanna get some untapped value out of your ChatGPT subscription, go into settings and turn on global dictation

It uses the same fast and accurate transcription path as the in-app dictation

[Movez @0xMovez]

this is f*cking crazy

karpathy's CLAUDE.md hit #1 on github trending.
228,000 stars. most devs still haven't read it.

it took my Claude agents accuracy from 65% to 94%.

the 6 rules inside:

• step 1 → surface assumptions

"State your assumptions explicitly. If uncertain, ask."
ambiguous task → list interpretations → ask → build.

• step 2 → cut the scope

"No abstractions for single-use code."
200 lines → refactor → 50 lines → ship.

• step 3 → keep the diff surgical

"Don't refactor things that aren't broken."
request → minimal diff. every changed line traces back.

• step 4 → clean only your orphans

"Remove imports/variables/functions that YOUR changes made unused."

old dead code → flag it, don't delete it.

• step 5 → turn tasks into tests

"Fix the bug" → "Write a test that reproduces it, then make it pass."

vague task → failing test → pass/fail signal.

• step 6 → run the verify loop

"[Step] → verify: [check]"
plan → code → verify → fail? loop → pass → ship.

that's it.
65 lines. 6 rules. 94% accuracy.

save this before everyone else does.

then, read the article on how to build a team of Claude agents with Graphs below ↓

[Vox @Voxyz_ai]

Vibecoders, 𝗜'𝗺 𝗯𝗲𝗴𝗴𝗶𝗻𝗴 𝘆𝗼𝘂, open your messiest project and paste this into Claude Code or Codex:

"Find all of the implicit state machines in my app. Make the best decision that will elevate all three: the user experience (UX), developer experience (DX), and agent experience (AX) always, without breaking anything."

[chuplung @choopyplug1]

Claude Code limits just got F*CKED by a solo dev...

he dropped a free plugin that makes your agent burn 5x fewer tokens

you're not hitting your limit because you code too much. your agent re-reads the same chat over and over

here's how to use and set up billion-context:

first understand how it works:

• every message, Claude Code sends your whole chat again. that's where your limit goes

• billion-context sits between Claude Code and the model and squeezes old messages into tiny summaries

• the model decides what to compress and can unzip it back when it needs it. nothing gets lost

here's how to set it up for FREE:

1 → open Claude Code in any project

2 → say: "install Node.js if I don't have it, then run npm install -g billion-context --prefix=~/.local and bili plugin install claude"

3 → restart Claude Code. compression is on

4 → check it works: curl -s http://localhost:8787/__bili/health

5 → work like normal. the first compression kicks in after about 50K tokens, you don't touch anything

6 → the step everyone skips: don't run it next to another compression plugin. it compresses twice and wrecks your context

7 → run it from the terminal. the Claude Desktop Code tab skips the plugin

want proof? 174,327 real model calls over 4.5 months. 58% cheaper than full context. 1.48M tokens squeezed 30x, where full context crashed 199 out of 199 times

works with Claude Code, Codex, Gemini CLI, Aider, Zed and 15+ more agents

repo link + my full guide on the other 9 limit leaks below ↓

[별똥별 @stellargent_lab]

GitHub 스타 약 1.2만 개. 바이브코딩으로 만든 AI 티 나는 화면을 디자이너가 손본 화면처럼 바꿔 주는 스킬이 있어요.

이름은 'unslop'.

보라색 그라데이션, 빛 번짐 효과, 이모지 라벨, 과하게 외치는 문구, 들쭉날쭉한 간격 같은 AI 티를 먼저 찾아냅니다. 그다음 제품에 맞는 방향을 정하고, 디자이너가 검토하듯 여러 번에 나눠 고쳐요.

Claude Code, Codex, Google Antigravity에서 쓸 수 있고, React나 Vue, 일반 HTML 등 지금 쓰는 방식 그대로 작업합니다. 링크만 있는 운영 중인 사이트도 다룰 수 있어요.

고치기 전에 진단 결과와 방향부터 보여 줘서 중간에 방향을 틀 수 있습니다. 이미 괜찮은 화면이면 갈아엎지 않고 작은 개선만 제안하고요.

이미지가 필요하면 유료 이미지 생성 서비스를 따로 연결해야 하고, 생성할 때마다 먼저 물어봅니다. 화면이 바뀌는 과정을 담은 10초 미만 영상도 만들 수 있는데, 이건 별도 도구 설치가 필요해요.

라이선스는 CC BY 4.0이라 출처만 밝히면 상업적으로도 쓸 수 있습니다.

AI로 만든 화면이 다 비슷해 보이는 게 바이브코딩의 문제점이라 보는데 AI티를 벗어나는게 필요하겠네요

[John Eliel Baladjay @JohnECBal]

One rule I had to write for my AI: "save this" does not mean "do this."

Without it, a note quietly turns into a task, a reminder or a promise I never made.

[Google Antigravity @antigravity]

Today, we are launching a new Knowledge Work interface layout tuned for research, writing, and artifacts: same agent harness, plugins, and MCP tools, without the terminal and git clutter. Select your layout in the General tab of Settings.

Our goal with Antigravity is to help the world deliver value through products and research. Doing that always starts with building the logic, which is why software development is at the core of the product. But turning that work into real outcomes takes every role working together with a shared language and understanding.

[John Eliel Baladjay @JohnECBal]

I don't set my AI's permissions by how big a task is. I set them by who sees the mistake.

It can update our business records on its own, because we can fix those inside the company. It can't send a single message until I've read the exact words and typed "send."