近期 App 要上线，周末过来 Vibe Coding 非常高效无人打扰，平时应酬太影响进度。

🧠 I think I’m close to the upper bound of human–AI collaboration efficiency. Not because of one tool — but because the system finally clicked. Here’s what I’m running right now: • Claude Code Agent Teams • Custom subagents • Strict rules / skills / commands • git worktree

🧠 我觉得自己已经接近了人类 × AI 协作效率的上限。 不是因为某一个工具有多强， 而是整个协作系统终于跑通了。 我现在使用的组合是： • Claude Code 的 Agent Teams • 自定义 子代理（subagents） • 严格的 rules / skills / commands • 用 git worktree 实现真正的并行开发 这已经不是“AI

“Trying to fully utilize the quota. Accidentally triggered anxiety.””😅

作为AI开发者，我最头疼的痛点：各大平台的配置页面居然没API或HTTP(S)接口！结果呢？天天手动写一堆 Commands 适配，时间全浪费了😩。谁有好方案？快来分享！#AIDev #Automation #Claude #VibeCoding

Kimi K2.5 + Seedance 2 is a perfect workflow. One prompt = A 100MB Excel storyboard generated with images and prompts, which you can use in seedance. I just tested with Hitchcock's Psycho iconic shower scene. My prompt: I would like to create a derivative work of the

AI is just stupid when it comes to understanding new knowledge. #ClaudeCode

Nano Banana Pro and GPT image 1.5 Prompt: Macro food photography of a crispy martabak telur tearing open mid-air, thin golden wrapper shattering with a crack, hot filling of minced beef, scallions, and egg bursting outward. Flecks of fried onion and pepper suspended in motion,

Everyone’s out here flexing their fancy AI prompts and tricks 🌟 I’m just good at turning AI into actual shipped products 🛠️

Happy New Year. Let’s build something great in 2026. 🚀

Already running Opus 4.6 in production for a while—it’s still the beast for the deepest agentic planning, massive codebases, and those ultra-long persistence tasks where every edge matters. Sonnet 4.6 looks seriously impressive though: closing the gap hard on computer use, coding

Wow—this per-slide revision capability in NotebookLM hits exactly the iterative sweet spot for source-grounded presentations. Being able to refine individual slides via targeted prompts (text adjustments, layout tweaks, visual updates) without regenerating the entire deck

App Store 驳回两次，打回处理倒是没什么审核周期也太 TM 长了接近两周🙁

最近 App 产品再做找音乐素材这不就来了嘛

收到第二次驳回邮件啥都没干立马写个 skills 扮演 App Store 评审专家走一遍流程

非常讨厌苹果的效率

Gemini 3.1 Pro just dropped

“As an AI builder, Gemini 3.1 Pro hitting 77.1% on ARC-AGI-2 has me buzzing. From ISS dashboards to atmospheric UIs—this thing gets systems. Time to build. 🔥”

Gemini 3 Pro vs Gemini 3.1 Pro