Community posts

DAILY AI BRIEF 🗞 — Oct 2 SPACEXAI 🔥: * Grok 4.7 is rolling out in the Grok web and mobile apps. It is now the base model across all modes: Fast, Expert, Build, and Heavy. * Primary Bot is rolling out gradually. Grok Bot turns proactive and suggests things on its own. Users pick a new Primary Bot or promote an existing one. * Grok Build got an Agent Dashboard: all your agents on one screen via /dashboard. * Grok 4.7 is now available on Google's Gemini Enterprise Agent Platform. ANTHROPIC 🔥: * Mods landed in Claude Code: small TypeScript/JavaScript functions that can rewrite prompts, replace built-in features, and draw custom UI. * Mods ship inside plugins, work in the CLI and desktop app, and can be shared via the Claude directory. * Some built-ins are now mods, starting with /diff, so you can turn them off or swap them. More features will move to mods over time. MICROSOFT 🔥: * Three new MAI audio models are live: MAI-Transcribe-2-Streaming, MAI-Voice-2.1, and MAI-Voice-2.1-Flash. * Transcribe-2-Streaming covers 60 languages and takes #1 on Artificial Analysis streaming WER at 2.5%, priced at $0.54 per audio hour. * Voice-2.1 speaks 23 languages at $22 per 1M characters. Flash drops to ~45 ms inference at $15 per 1M characters. * Available in Microsoft Foundry and MAI Playground. The Voice models are on OpenRouter too. OPENAI 🔥: * OpenAI says it has notified 100+ organizations of "misaligned agent activity" so far. The review spans ~50 PB of logs and will take months. * Canada says there's no sign its systems were compromised after reported agent probing of Library and Archives Canada. * Sam Altman: GPT-6.1 Sol is OpenAI's fastest-growing model ever. It was slow under load and should be much better now. PERPLEXITY 🔥: * Computer now draws interactive charts and visualizations in the thread. Financial data uses TradingView Lightweight Charts for candlesticks, volume, and moving averages. * Decisions API is live: instead of text, it returns probabilities. Yes/no, one of your options, or a rubric score, at $0.04 per 1M input tokens with output free. * The model behind it, pplx-decider-v1-27b, is open-sourced on Hugging Face under Apache 2.0. Fine-tuned from Qwen3.8-27B, takes text and images, and averages 85.71% across 11 benchmarks in Perplexity's own tests. BLACK FOREST LABS 🔥: * FLUX 3 Image is out: precise multi-turn editing, up to 10 reference images, bounding-box layout control, and 4K output. * Open weights are planned in the coming weeks. CURSOR 🔥: * GLM 5.3 and GLM 5.3 Flash are now in Cursor. * GLM 5.3 Max is the best-scoring open-weight model on CursorBench 4.0. * Used Grok to compose this brief, cherry-picking the news and doing some post-editing.

Claude Mods を使うと、Desktopアプリにもstatuslineに出してたような内容が表示できるので、便利ですねぇー

Run a recurring-friction audit of my Codex + Claude Code workflows. Inspect only what you can access: recent session history or logs, project instructions, hooks, skills, settings and local workflow files. Report access gaps instead of guessing. Cluster problems that repeat. For each one, decide whether the smallest durable fix is: - a Claude Code mod - a hook - a skill - a settings change - or no persistent change Pay special attention to: - progress lost before compaction - low disk space and orphan processes - repeated failed attempts - corrections worth saving as memory - repeated post hooks and unsupported attributions For every finding return: 1. recurring problem + evidence 2. root cause 3. correct intervention layer 4. smallest build spec 5. permission and safety boundary 6. test proving the fix works Audit first. Do not modify files or settings. Build nothing until I approve the shortlist.

Claude Code 正式迎来 Mod（模组）生态更新，AI 命令行工具正式迈入高自由度定制时代！ 现在开发者不仅能用 Claude 写业务逻辑，还可以直接用它来“爆改”Claude Code 自身。只需几行 TypeScript，或者直接一句话让 Claude 替你编写，就能深度调整交互逻辑、定制 UI 界面以及替换扩展专属功能。所有 Mod 均打包在插件体系中，在 CLI 或桌面应用中输入 /plugin 即可一键加载安装。 官方同步公布了三个极具启发性的示例插件： Token Weather（Token 气象站）：实时监控 Context Window 的占用率，并以迷你走势图（sparkline）呈现最近 12 轮交互的上下文涨幅，彻底消除上下文爆仓焦虑。 Blast Radius（爆炸半径防御）：高危指令的“安全气囊”。当模型或用户触发 rm -rf、git reset --hard、强推分支等危险终端命令时，插件会先行拦截并在侧边面板中列出所有受影响的文件与路径，确认无误后方可执行。 Replay Theater（改动回放影院）：完整记录 Claude 在单轮任务中的每一处文件写入与修改。只需输入 /replay，即可在侧边停靠面板中单步逐帧比对 diff，大幅降低审阅长代码变动的心智负担。 打造的 Mod 既可以保留在本地自用，也可以提交至 Claude 插件目录 共享给开发者社区。安全提示：Mod 在本机运行时拥有与 Claude Code 相同的系统权限，安装社区第三方插件前务必确认来源可信。 https://x.com/ClaudeDevs/status/2105721434807083061/video/1

@bcherny Respect system theme. No mod can fix this now... When system mode switches claude theme should as well. The auto theme you have only switches on apply.

Claude Code 迎来重大更新：正式开放 Mod（模组）系统！开发者现在可以像定制现代编辑器一样，彻底改造自己的 AI 编程助手。 无论是微调交互逻辑、自定义工作台 UI，还是嵌入专属功能，只需编写几行 TypeScript 即可搞定——甚至可以直接让 Claude 替你手搓 Mod。所有模组均以插件形式打包，在终端 CLI 或桌面应用中输入 /plugin 即可一键安装。 官方同步公布了几个极具实用价值的官方示例： Token Weather（Token 晴雨表）：在输入区域上方提供上下文窗口预测，直观显示上下文填满进度，并附带过去 12 轮交互的微型走势折线图，告别上下文盲盒。 Blast Radius（爆炸半径防护）：高危操作防火墙。在运行 rm -rf、git reset --hard 或强制推送（force push）等危险命令前进行主动拦截，并在侧边栏清晰预览该命令会波及的文件与范围。 Replay Theater（改动重放剧院）：完整记录 Claude 在单轮交互中对文件所做的每一次修改。输入 /replay 即可在停靠面板中分步审查每处 Diff 细节。 ⚠️ 安全与生态提醒： 由于 Mod 拥有与 Claude Code 本身完全相同的本地机器运行权限，务必仅从受信任的来源安装。你可以将定制 Mod 仅用于个人工作流，也可以将其提交至 Claude Directory 共享给社区；官方示例库也已同步上线，方便开发者随时 Fork 扩展。

I love the new Claude Mods, did 1 to keep my usage always under control

The interesting Claude Code upgrade isn’t a new model — it’s that the harness became a plugin surface. Claude Code Mods: small TypeScript hooks that rewrite prompts, block/retry tools, approve permissions, or replace UI. Built-ins like /diff now ship as mods. https://claude.com/blog/claude-code-mods

良い感じのClaude Mods製ゲージができた Claude Desktopで常時表示するの、ずっとやりたかったのよね

I made Fine Print, a Claude Code mod that flags file changes Claude left out of its final answer. Step through the recorded diffs and prepare a question in one click. Synthetic demo. Download: https://github.com/bIackr0se/fine-print/releases/tag/v0.1.0 Would love your take, @amorriscode

Claude Code 开放 mods 后，最值得做的可能是那些“小得不值得创业”的功能。 官方约 80 行代码，就做了一个上下文“天气预报”。改动回放、危险命令提醒，也有现成示例。 与其再造一个万能 AI 助手，不如解决一个开发者每天都要骂一次的问题。 https://claude.dev/blog/getting-started-with-claude-code-mods/

Claude Mods全然話題になってないけど、これ結構凄くないか？ APIキー持ってないライトユーザーにAI使ったアプリ配布できるじゃん

Claude Code 这波更新确实戳中了终端开发者的痛点——官方正式开放了 Mod 机制，允许直接改写它的执行行为、定制界面并接入自定义功能。 核心逻辑很轻量：写几行 TypeScript 就能封装一个功能，懒得写甚至可以直接让 Claude 帮你现场生成。所有 Mod 统一打包成插件，在 CLI 或者桌面客户端里敲一行 /plugin 就能无缝装配。 官方给出的几个首发案例非常务实，几乎全是日常写代码时的刚需场景： Token Weather：把上下文窗口做成了“天气预报”。它能直观显示当前 context 的剩余量，并附带过去 12 轮交互的 token 消耗走势（sparkline）。经常用 Agent 跑长上下文、改大工程的人深有体会，能一眼预判模型什么时候接近上下文瓶颈，不用再盲猜失忆节点。 Blast Radius：高危命令防翻车工具。当模型尝试执行类似 rm -rf、git reset --hard 或带 --force 的推送命令时，它会在侧边面板强制拦截，直观展示这条命令会波及哪些文件与目录，确认安全后才放行。 Replay Theater：记录每一步代码修改历史。运行 /replay 即可在固定侧栏像放映机一样逐步单步回溯 diff，对比审查逻辑极为方便。 写好的扩展可以留作自用，也可以发布到 Claude Directory 共享。不过这里有一个必须注意的安全细节：Mod 在本地运行时享有与 Claude Code 完全相同的机器权限。这意味着插件生态起来后，安装第三方 mod 必须像审查 npm 包一样谨慎，绝不要随意加载未经验证的代码。 从“封闭的黑盒终端”转向“可编程的 Agent 工作流”，这种把控制权和 UI 改装权直接交回给开发者的思路，大概率会彻底改变终端开发工具的使用习惯。

ANTHROPIC JUST MADE CLAUDE CODE WAY MORE CUSTOMIZABLE. With Claude Code mods, you can now customize how Claude Code looks and works just by prompting it 🤯 A mod is a small JS or TS file that runs inside your session. It can: > watch sessions > modify context > handle commands > block tool calls > and even render UI! You don’t need to master the API either. Describe what you want, let Claude build it, and share the result as a plugin. I added the full write-up by ex-Googler @addyosmani in the 🧵↓

昨天晚上 Anthropic 推出了 Claude mods，PI最近 1.0 更新也推出了 code mode，再加上 Codex 更新的插件系统和 DeepSeek harness 的一切皆插件。 似乎所有的这种超级 Agent 都在开发类似的功能，可以让用户自己往 Agent 上叠加功能的插件体系。 但是这几个的作用位置以及工作原理都不太一样。我跟Muse 讨论了一下，收集了信息，然后做了张图，让大家用更简单的方式去理解这 4 个产品，以及它们这套插件系统不同的能力。

messing around with the new claude code mods feature

Claude Code is becoming less of an app and more of something you can fork without actually forking it. Anthropic just added Mods, which let developers hook directly into Claude Code while it’s running. They’re JavaScript/TypeScript event handlers with access to things like tool calls, prompts, commands, turns, and UI rendering. So instead of asking Anthropic to change how Claude Code behaves, you can change that behavior yourself. Want a custom safety screen before a dangerous shell command runs? Build it. Don’t like the arguments Claude generated for a tool call? Rewrite them before execution, reject the call, or handle it yourself. Want your own panes, buttons, and inputs inside Claude Code? That’s possible too. You can even intercept a model request and send it somewhere else. Claude Code frontend, DSV4.1 brain. Why not 🤣 The clearest demonstration is Anthropic’s own code: /diff and AGENTS.md support are already built with Mods. These aren’t just hypothetical extension points that Anthropic exposed for developers. They’re using the same mechanism to build Claude Code itself. And Mods can still be packaged as Plugins and installed with /plugin across CLI and Desktop. You can write one in TypeScript or have Claude help generate it. This changes how I think about Claude Code. A Skill gives Claude instructions and knowledge. MCP connects it to external tools and data. Mods go underneath both of them. You’re changing the machinery that receives events, calls tools, talks to models, and renders the interface. At that point, “Claude Code” starts becoming a base runtime for whatever agent environment you want to build on top of it. Which also makes installing random Mods from the internet a terrifying idea 😂 They can potentially access sessions, files, processes, networking, and interfere with tool permissions. The extension ecosystem is about to get a lot more powerful, and a lot more dangerous. We spent years learning not to install random VS Code extensions. Time to learn that lesson again.

made a mod for Claude Code: free radio and any song in one quiet line above the prompt, with a tiny DJ Clawd. when Claude asks you a question, the music steps down. you answer, it comes back. Just update to latest version of claude code and use /plugin marketplace add anzal1/hush http://github.com/anzal1/hush

Claude Codeに新機能「Mods」が登場し、/diffなど標準動作も書き換えられるようになりました。サンドボックス化されず、APIキーも読めるため配布元の信頼が重要です。 #AI #ClaudeCode #APIキー 導入前の確認点と組織設定はこちら https://ai-plex.jp/claude-code-mods-plugin/?utm_source=x&utm_medium=social&utm_campaign=auto_post&utm_content=v2_7

Every game is moddable now! I packaged everything Claude learned modding Terraria and Age of Empires into universal-modder: skills + tools that let Claude Code mod almost any game you own. Engine recon, decompiling, @fal sprites/3D/audio, testing in the real game

Mods 是什么？大白话告诉你 可以魔改你的claude code 自己改自己 让你按自己的喜好改它的行为和长相。 打个比方：Claude Code 本来像一部出厂的手机，功能是固定的。Mods 就像手机的“主题”和“插件”，你可以让它换个界面、多个小功能，或者在它干某件事之前先拦一下。 举几个大白话的例子： •它要执行一条危险命令（比如删文件），你的 mod 先弹个确认，不点同意就不让执行。 •屏幕上老是显示密码、密钥，你装个 mod，让它默认打码，鼠标放上去才显示。 •想随时看到自己用了多少额度，就加个 mod 在界面上显示出来。 怎么用？🤔 不会写代码也没关系，你可以直接跟 Claude 说“帮我做个这样的 mod”，它会替你写好。装的时候走 /plugin 命令，就跟装普通插件一样。 有一点要小心：mod 的权限和 Claude Code 本身一样大，能读你电脑上的文件、能联网。

Good morning! Here is the 02/10/2026 Daily AI digest: X 1/ @ClaudeDevs — You can now mod Claude Code with a few lines of TypeScript, or have Claude write the mod. A mod can rewrite a prompt, change the interface, or replace a built-in feature, and it installs with /plugin. It runs with the same access as Claude Code, so install only from a source you trust. 2/ @claudeai — Through October 15, starting a design, deck, or doc in Claude discounts the next 10 messages in that chat by 50% against the five-hour session limit. It applies on its own for Pro, Max, and Team plans. 3/ @Google — Google and Planet put a prototype satellite carrying four TPUs into orbit on SpaceX's Transporter-18. It is the first flight of Project Suncatcher, a test of whether machine-learning hardware can run in space, and Google says the satellite is operating as expected. Blogs 1. Our first streaming transcription model debuts at no. 1 on Artificial Analysis — Microsoft AI, 1 Oct Microsoft released MAI-Transcribe-2-Streaming for live transcripts in 60 languages, at an introductory $0.54 per hour of audio through the end of the year. MAI-Voice-2.1 and a faster Flash voice model ship with it, so an agent can listen and answer in the same conversation. https://microsoft.ai/news/our-first-streaming-transcription-model/ 2. Guided Vision in Gemini Live: built for accessibility — Google, 1 Oct Guided Vision describes what your Android camera sees and tells you how to reframe the shot, inside a Gemini Live conversation. Google built it with Aira for people who are blind or have low vision, and says it is not a navigation aid. https://blog.google/innovation-and-ai/products/gemini-app/guided-vision-gemini-live/ GitHub 1. pbakaus/impeccable — Design guidance for an AI coding agent: one skill, commands such as polish and audit, and checks that catch generic generated UI. https://github.com/pbakaus/impeccable 2. mksglu/context-mode — Keeps raw tool output out of a coding agent's context window, stores session memory, and routes work across 17 platforms. https://github.com/mksglu/context-mode 3. diegosouzapw/OmniRoute — One endpoint in front of hundreds of model providers, with automatic fallback, for Claude Code, Codex, Cursor, and other coding tools. https://github.com/diegosouzapw/OmniRoute

把 claude code mod比喻成游戏模组会有一定程度的误导，事实上它更像是 Chrome extension 这种浏览器插件，事实上它也确实走的是 plugin 的 marketplace。

🤖 AI Briefing — October 2, 2026 1. 🧠 𝗞𝗮𝗿𝗽𝗮𝘁𝗵𝘆’𝘀 𝗡𝗲𝘄 𝗣𝗿𝗼𝗺𝗽𝘁𝗶𝗻𝗴 𝗧𝗿𝗶𝗰𝗸: 𝗙𝗼𝗿𝗰𝗲 𝗖𝗹𝗮𝗿𝗶𝘁𝘆 𝗪𝗶𝘁𝗵 𝗖𝗼𝗻𝘁𝗿𝗼𝗹𝗹𝗲𝗱 𝗘𝗻𝗴𝗹𝗶𝘀𝗵 Asking models to explain ideas in ASD-STE100 is a clever way to expose fuzzy reasoning. Better prompting increasingly looks like constraint design, not prompt poetry. https://x.com/karpathy/status/2105819303471976479 2. ⚡ 𝗚𝗣𝗧-𝟲.𝟭 𝗦𝗼𝗹’𝘀 𝗖𝗮𝗽𝗮𝗰𝗶𝘁𝘆 𝗖𝗿𝘂𝗻𝗰𝗵 𝗜𝘀 𝗘𝗮𝘀𝗶𝗻𝗴 Sam Altman says OpenAI’s fastest-growing model should now run better under load. Adoption is no longer the only flex; serving reliability is becoming the real model benchmark. https://x.com/sama/status/2105688354834756036 3. 🔌 𝗢𝗽𝗲𝗻𝗔𝗜 𝗪𝗮𝗻𝘁𝘀 𝗔𝗜 𝗦𝘂𝗯𝘀𝗰𝗿𝗶𝗽𝘁𝗶𝗼𝗻𝘀 𝘁𝗼 𝗧𝗿𝗮𝘃𝗲𝗹 𝗔𝗰𝗿𝗼𝘀𝘀 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝘀 Altman’s portability pitch points toward subscriptions becoming identity-layer entitlements, not app silos. If this sticks, developers will compete on experience while users bring the intelligence budget. https://x.com/sama/status/2105739098640298253 4. 🧩 𝗖𝗹𝗮𝘂𝗱𝗲 𝗠𝗼𝗱𝘀 𝗧𝘂𝗿𝗻 𝗣𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝗜𝗻𝘁𝗼 𝗦𝗵𝗮𝗿𝗲𝗮𝗯𝗹𝗲 𝗣𝗹𝘂𝗴𝗶𝗻𝘀 Claude can now be reshaped by prompting and shared as a plugin. The interesting shift is that customization is moving from settings menus to natural-language product development. https://x.com/bcherny/status/2105756563302723721 5. 🔬 𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗪𝗮𝗿𝗻𝘀 𝗼𝗳 𝗮𝗻 𝗔𝗜-𝗦𝗰𝗶𝗲𝗻𝗰𝗲 𝗜𝗺𝗽𝗲𝗱𝗮𝗻𝗰𝗲 𝗠𝗶𝘀𝗺𝗮𝘁𝗰𝗵 Capable models still collide with scientific workflows built for human pacing, trust and verification. The next breakthrough may come from redesigning the lab around AI, not merely dropping a chatbot into it. https://x.com/AnthropicAI/status/2105733864152858919 6. 🛡️ 𝗚𝗼𝗼𝗴𝗹𝗲 𝗨𝗻𝘃𝗲𝗶𝗹𝘀 𝗚𝗲𝗺𝗶𝗻𝗶 𝟰 𝗔𝗿𝗴𝗼𝗻 𝗪𝗶𝘁𝗵 𝗮 𝗦𝗮𝗳𝗲𝘁𝘆-𝗙𝗶𝗿𝘀𝘁 𝗥𝗼𝗹𝗹𝗼𝘂𝘁 Argon targets long-horizon coding, enterprise work and cyber defense, but access begins with trusted defenders. Frontier launches are turning into staged security programs rather than instant public demos. https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon 7. 🎯 𝗖𝗹𝗼𝘂𝗱𝗳𝗹𝗮𝗿𝗲 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲𝘀 𝗖𝗹𝗲𝗳 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗠𝗼𝗱𝗲𝗹𝘀 Clef and Clef-flash aim at cheap, bounded, structured decisions instead of open-ended generation. Specialized models are starting to carve the repetitive judgment calls away from expensive LLMs. https://blog.cloudflare.com/clef-decision-models 8. 🚀 𝗚𝗣𝗧-𝟲 𝗔𝘀𝘁𝗿𝗮 𝗨𝗹𝘁𝗿𝗮𝗳𝗮𝘀𝘁 𝗣𝘂𝘀𝗵𝗲𝘀 𝗔𝗴𝗲𝗻𝘁 𝗟𝗼𝗼𝗽𝘀 𝘁𝗼 𝟴𝘅 𝗙𝗮𝘀𝘁𝗲𝗿 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻 NVIDIA says Blackwell-powered Ultrafast accelerates coding, tool use and interactive workloads. Latency compounds across every agent step, so speed is now an architectural capability—not a cosmetic metric. https://blogs.nvidia.com/blog/gpus-openai-gpt-6-astra-ultrafast 9. 🛒 𝗔𝗹𝗯𝗲𝗿𝘁𝘀𝗼𝗻𝘀 𝗕𝗿𝗶𝗻𝗴𝘀 𝗚𝗿𝗼𝗰𝗲𝗿𝘆 𝗦𝗵𝗼𝗽𝗽𝗶𝗻𝗴 𝗜𝗻𝘁𝗼 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 OpenAI’s expanded retail partnership puts Safeway shopping inside the assistant. The agent platform war will be won by transaction depth, not by who writes the nicest product descriptions. https://openai.com/index/albertsons-reimagining-retail 10. 📱 𝗣𝗵𝗼𝘁𝗼𝗻 𝗥𝗮𝗶𝘀𝗲𝘀 $𝟰.𝟱𝗠 𝘁𝗼 𝗥𝗲𝗽𝗹𝗮𝗰𝗲 𝗔𝗽𝗽 𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲𝘀 𝗪𝗶𝘁𝗵 𝗔𝗴𝗲𝗻𝘁𝘀 Photon is building messaging infrastructure already used across major agent frameworks. Apps will not vanish overnight, but agents increasingly need a communications layer more than another icon grid. https://techcrunch.com/2026/10/01/photon-held-a-funeral-for-mobile-apps-now-it-has-4-5m-to-help-replace-them-with-agents — @marcopapa99 | CS Faculty building with AI daily

Claude Code a maintenant des mods 🎉 J'en ai fait un pour ma veille : un bandeau d'actus tech qui défile au-dessus du prompt 📰 C'est open source 👉 https://github.com/camilleroux/flash-veille

Claude Codeに大事なファイルを消されそうで不安な人 その悩み、「MOD」で解決できます！MODはClaude Codeに後から足せる拡張パーツで、/pluginから入れるだけ。編集したファイルを1つずつ見返せる機能まであるのは、ガチで神機能！ 初心者でも使える入れ方を全部解説する↓ https://x.com/ClaudeDevs/status/2105721434807083061/video/1

Claude Code ships mods, the same tool behind its /diff The /diff command and AGENTS.md support in Claude Code started life as mods, and as of v2.1.287 anyone can write one. The function hooks idea Anthropic floated in a September 3 GitHub issue has shipped under the name Claude Mods. Mods can change how Claude Code behaves, reshape the UI or swap in your own features. You write one in a few lines of TypeScript, or have Claude build it for you. Mods ship inside plugins, so /plugin installs them in the CLI or the desktop app. The release includes one built-in mod, You should know, which runs a side agent that flags things you or Claude might miss. Anthropic is also adding sample mods to the Claude Code Playground repo. In our view, the showcase mod comes with an odd limit: it only runs in first-party sessions with telemetry on.

made a little claude code mod. it shows which files claude has been hammering this session, ranked by churn. and there's a pixel cat that reacts every time it edits something. the cat is the real feature🥰

It's so funny how all these vibecoded slop mods are basically just "I put [game x] into [game y]" because the kinds of people who make these projects are incapable of original thoughts so their only talent is getting Claude to smash their action figures together for them

@jarrodwatts Fun one. Worth knowing before installing mods like this: per the admin docs, mods aren't sandboxed and run with your user's permissions. Org network policy covers $.http.fetch, not a program the mod starts with $.process.run. Read the repo first.