Live mentions
Real-time mentions from X plus the latest from Reddit and Hacker News. Updated every 10 minutes
Live mentions
Real-time mentions from X plus the latest from Reddit and Hacker News. Updated every 10 minutes
Chinese words and radicals on your iPhone Lock Screen, Home Screen and Mac desktop
I’m learning Chinese and saw some apps that show words on the phone screen without open app Based on that idea, i made 2 iPhone widgets: Lock Screen: a new word every few minutes, you can tap it to…
Your AI agent needs your API keys. It doesn't need to see them. tokenstash writes keys straight into .env.local for Claude Code, Codex, Cursor and Gemini CLI. Paste each key once, in your browser, never in the chat.
keep your API keys out of chat and let your agents keep building. introducing tokenstash: the keychain for coding agents. save a key once. approve a project and tokenstash drops it into .env.local.…
37623
A live map beside the chat that shows what Claude Code is doing: tool calls, model steps, hooks, skills and subagents light up as they happen. Local only, in five languages.
I kept losing track of Claude Code runs: subagents still going in the background, a hook refusing calls, context creeping up. Tried lot-map, a live map beside the chat. Calls pulse to stations…
$TRUE: a token mined from what a market believes and burned when it stops. NEAR Hackathon 2026.
I also built $TRUE at NEAR HACKATHON: a token whose supply is mined from what a market believes, and burned when it stops. Bond NEAR to a claim, mine $TRUE while it holds. Refuted? The tokens burn…
1.4k49
Entity-level git merge driver. Resolves false conflicts git invents when independent agents edit the same file. ~95% reduction vs. line-based merge.
your agents are done. your ci isn't. git + ci reruns all 13,742 tests on every push and merges by line. agents wait 18:48 to hear back. land reruns only the tests that read the change, and merges by…
3.5k135
Semantic version control => entity-level diffs, blame, and impact analysis on top of git. 28 languages via tree-sitter. Built for coding agents.
your agents are done. your ci isn't. git + ci reruns all 13,742 tests on every push and merges by line. agents wait 18:48 to hear back. land reruns only the tests that read the change, and merges by…
134190
Mike's own agent app: a runner daemon plus a native Mac app
Things that just make sense in 🧵🐻 #6: Starred threads stay at the top, even while they work The ones I really care about don't disappear off into Running when I kick them off, they just stay right…
60 bilingual prompts and an agent skill for clear, evidence-grounded writing across research, work and everyday communication.
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
Anti defensive-writing. Modern LLMs (especially GPT) often generate overly defensive outputs. This skill helps counter that tendency.
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
Remove unrequested caveats, faux rigor, invented metrics, and AI formatting residue from academic writing.
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
公开演示版：识别并修复中文社科论文中的防御性写作，不虚构事实、数据与文献。
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
极致轻量｜阻止论文的防御性写作 | Anti-defensive writing for academic papers — 学术写作原则 Skill + prompt pack (CN/EN)
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
Codex skill for removing defensive writing and strengthening prose.
想骂个软件难用，先夸它有价值；想嫌 AI 太贵，先承认它能提效——现在的文章和 AI 越来越不敢直接说人话了，满屏都是怕被杠的「防御性写作」和怕得罪人的「防越界写作」。怎么把这些废话一刀砍掉？下面整理了 6 个现成的开源 Skill、2 篇论文结论和一段直接能抄的提示词 👇
Chat with Claude Code on Discord. Support channel modes (free response or mention-only), threaded messages as claude sessions, allowlist, scheduled watcher jobs
Every morning I wake up to reports in Discord. 2 AM: my AI drafts a blog article 3:35 AM: 4 fresh video ideas 4 AM: it cleans up the server's disk Every 2 min: it checks my GitHub comments and starts…
38.8k10.5k
SGLang is a high-performance serving framework for large language models and multimodal models.
Our Q4 multi-hardware roadmap is out. SGLang now runs on 10+ accelerator families, and the plugin system makes adding a new backend out-of-tree a first-class path. In-tree backends, Q4 focus: 🔸…
Run a Flutter app on multiple device sizes side by side in one macOS window, with mirrored input.
複数の端末サイズUIを同時に確認するようのFlutterパッケージちょっと前に作った👀 multi_device_preview
I built an Android AI agent that uses your phone for you. One sentence, and it opens the app, taps through, and finishes the task on a hidden background screen while you keep scrolling. Mid-task a…
Measured CapyCTL deployment recipes per engine, model and GPU
Me too: ~1,560 tok/s combined with 8 lanes, Qwen3.8-27B NVFP4 + DFlash2, on the released @vllm_project 0.32 through CapyCTL. Caveat: an RTX PRO 6000 (same chip and bandwidth as a 5090) with a 16GiB…
43.2k6.5k
The self-driving agent platform
Fixed a shell-output decoding bug in @agno — run_command failed on non-UTF-8 bytes while arun_command returned them fine. One shared decoder now keeps both in sync across every locale. #OpenSource…
118.3k10.3k
This project is dedicated to collecting high-quality macOS software and organizing them systematically by different categories for easy search and use.
Awesome Mac 2个月暴增 5k 🌟，现在 118k了，感谢 🙏 🫶
Payables is an AI accounts-payable agent that pays supplier invoice s in USDC on NEAR
You can try to break it. The red team page gives you the compromised agent's key: swap a payout account, hide a prompt injection, resend a paid invoice, or sign a payment to yourself. The contract…
Open-source framework for auditing reward hacking in AI agents
4/5 The project explores: • Adversarial scenarios and payload transformations • Per-attempt canaries • Observable-behavior oracles • Replay and structured reports • Statistical summaries and CI…
262.7k64.8k
The agent that grows with you
I kept typing /usage to check my Claude limit and still hit it mid-task. So I built usage-watch for @NousResearch Hermes Agent: every few turns it posts one line with how much of your session and…
30944
gliff for @omarchy now reaches machines that don't run gliff: RDP, VNC and RustDesk, a click away in the address bar. Log in once, get its desktop in a tab, or push past the screen edge and drive it…
34k3.3k
Community plugin to control Blender 3D with any LLM of your choice. Not affiliated with the official Blender Foundation.
3️⃣ three.js（11.8 万⭐） 在浏览器里做 3D 场景 4️⃣ GSAP（3 万⭐） 管缓动和文字动效，画面高级感靠它 5️⃣ Manim（9.6 万⭐） 3Blue1Brown 做数学视频用的动画引擎，讲解类内容首选