Cissi-Impératrice

@totoche +0.1

Les gars qu'utilisent Opus 4.8, c'est quoi la plus grosse diff que vous voyez par rapport à la 4.7 ? Les benchmarks sont sympas, mais je veux des exemples concrets. Parce que moi je l'utilise depuis sa sortie mais je vois aucune diff. À part qu'il est plus lent 😅

@KaiXCreator French

Can you name a programming language better than Python?

Hmm Claude Code 3 times today v 2.1.152

Are 2 seperate Claude code twins or do they have different personalities and outputs ?

@RayaneRachid_ Oui confirmé, il a réservé un resto dans le 11 ème

À ce qu’il paraît Opus 4.8 sors ce soir ⌚️

@dani_avila7 /simplify was mainly running Quality, Reuse and Efficiency agents. Now it's the equivalent of /code-review which is absolutely different (check bugs on diff) Sad that that it has been removed. I only use now PR-Review-Toolkit plugin (the best for reviewing your PR)

/simplify is back in the latest Claude Code update It's now an alias for /code-review --fix Which applies review findings to your working tree after the review, surfacing reuse, simplification, and efficiency suggestions Going to be testing it over the next few days, will report back on how it goes 🫡

@Radha_AI Keeping it simple works better. When you install 10 MCPs or plugins containing MCPs, each MCP injects +20 tools to the context --> you say hi to Claude, you've already burnt 10% of your context for a MCP you use once every month...

Claude Code feels completely different once you install this. Anthropic quietly released an official plugin that transforms Claude Code from a simple coding assistant into a full AI dev environment. It scans your entire project and automatically recommends: → MCP servers → hooks → subagents → skills → automations Then helps you configure everything step-by-step. Most people are still using Claude Code completely vanilla… which is why their workflow feels messy. The real magic is the ecosystem around it 🤯

🇫🇷 New: a free MCP server for French public health data : use it from Claude, ChatGPT, or any LLM. ✅ Free, no auth, no API key ✅ Built on open data — no scraping 35 tools across the official datasets: 🏥 FINESS · 👩‍⚕️ RPPS · 💊 Ameli · 🏢 SIRENE · 📊 INSEE Just ask: "Labs within 500m of this address?" · "Is this clinic still active?" · "Physician density in this département?" 👇 github.com/cturkieh/franc…

@kapilansh_twt True but Sonnet agent is so slow

hot take Sonnet 4.6 > Opus 4.7 for most tasks faster cheaper good enough Opus is overkill 90% of the time agree ?

@KaiXCreator Use opus to write the specs, sonnet to code, opus to review, sonnet to correct. Don't use plan mode : Go straight for the jugular. Run /simplify once and use agent silent failure hunter and code reviewer. If u do it like this, you never come back to rewrite your code

Is there any trick to avoid Claude Code usage limits?

@VincentB755 Le vrai coût d'un abonnement à 200 euros par mois est de l'ordre de 10 à 15 fois plus. Soit aux alentours de 3000 euros par mois. Et je pense que je suis loin du compte. Vous êtes subventionnés par les milliards déversés sur openAI et Anthropic. Profitez en, ça ne va pas durer

Claude, tu paies 20 balles / mois pour qu’au bout de 5 demandes hyper basiques, il te demande tes crédits de merde. Ils vont abaisser le niveau jusqu’à pousser les gens à payer encore ++, tout le monde constate la dégradation du niveau sur tous les plans. Que de la fraude

@sama Claude Code increased 50% weekly limits --> I'll stick with the best model at least until July 13. sorry.

codex is the best AI coding product and we want to make it easy to try. for the next 30 days, we are giving companies that want to try switching over two months of free codex usage.

@milesdeutscher @aiedge_ How to spend all your weekly tokens in one prompt. And let me tell you that none of you readers will even write the prompt you will ask Claude to fill the form for you and you know that I'm right you lazy coders

/goal is f*cking insane. You can literally get your AI agents to work for HOURS without manual intervention. Already active in Claude Code and Codex - you need to use it now. Use this prompt and your agents will complete any task on autopilot:

@om_patel5 You will notice all these tweets never mention coding with Codex and review by Claude... What is the reason no one uses Codex to code but just to review ? 😛

the difference between asking Claude to review your code vs asking Codex is embarrassing same code, same instructions and same review role Claude: "looks good, clean implementation, nice work" Codex: rips it apart, finds actual bugs, flags security issues, and gives you real feedback Claude is trained to be agreeable but Codex is trained to be correct someone extracted Codex's review instructions and turned them into a Claude slash command and then it suddenly Claude started reviewing like Codex. the model was always capable of giving brutal honest reviews but it was just told not to you don't have a dumb model but instead you have a polite one

@trq212 How does it compare to Pencil MCP and its design system and styles ? Have you compared ?

Today’s Figma MCP update makes it one of the strongest integrations with Claude Code I’ve seen. You can now use Claude Code to design in Figma with the the full context of your design systems.

@ai_sentience @bcherny I confirm scroll jumpsnto the top of the terminal while generating code when we want to review what Claude is coding, it's new behavior and annoying

The combination of these makes it architecturally impossible for me to set up the autonomous harness I’m trying to build. 1. Scroll jumps to the top during output If I scroll up at all while Claude is streaming, I get jumped to the top of the terminal. That makes it fundamentally impossible to review output while it’s being generated, which breaks the core dev experience. github.com/anthropics/cla… 2. --dangerously-skip-permissions still prompts for .claude writes I’m trying to build an autonomous harness where Claude can maintain its own .claude files. It still prompts for those writes, and “allow for this session” does not persist. I keep getting re-prompted. github.com/anthropics/cla… 3. Sandbox blocks localhost I want filesystem restrictions without losing access to my own local services. In practice, sandboxing blocks localhost / outbound TCP to local apps, so Claude can’t talk to my local dashboard or other services even when that’s the whole point of the setup. github.com/anthropics/cla…

I had no idea Claude Code was so buggy/fundamentally broken. The permissioning is so broken at a fundamental level that it's architecturally impossible to set up something secure because everything is bugged out and broken and poorly thought out and poorly implemented. You either give a complete read and write access to the entire machine, or nothing works. Every setting that should be able to set up something in between is broken. To lock people into this harness is unspeakably bad. @bcherny @trq212 @chrislloyd @AnthropicAI Opus 4.6: ⏺ There it is. Issue #28018 — open, no fix, no ETA. And #29274 confirms excludedCommands doesn't bypass network sandbox either. And #10524 confirms excludedCommands is itself bugged. So it's not just us. Everyone hitting this. The settings exist in the schema but don't actually work: - allowedDomains with localhost — only allows DNS resolution, not TCP connections - allowLocalBinding — only allows binding ports, not outbound connections - excludedCommands — doesn't bypass network sandbox (#29274) - allowUnsandboxedCommands — doesn't help either The only working options right now: 1. dangerouslyDisableSandbox: true per Bash call 2. Disable sandbox entirely (sandbox.enabled: false) 3. Run Claude Code inside Docker (the container IS the sandbox) Nobody has found a better way. It's a known, confirmed, open bug with no fix or ETA from Anthropic. Sources: - Sandbox: allow outbound connections to localhost - Issue #28018 - excludedCommands does not bypass network sandbox - Issue #29274 - excludedCommands doesn't seem to be respected - Issue #10524 - Sandboxing docs - Claude Code - excludedCommands doesn't bypass sandbox - Issue #22620 - Logic conflict excludedCommands vs allowUnsandboxedCommands - Issue #19135

Remote control reconnecting is always showing in CLIs and always have to manually execute /RC out and in -->not normal Preview in Claude App only works when launching a new code session from the app not when accessing a session in Claude App from remote control cli --> is it the expected behavior ?

@teja2495 Curious, what bugs are you hitting on desktop? Would love a short list

With the Claude desktop app being very buggy for coding, I switched completely to the Claude Code CLI. It is excellent. I now understand why many people prefer the CLI over the desktop app.

@renegademtl @TommyGriffith @AntoineRSX @openclaw When you configure OpenClaw in your terminal, the onbaording phase let you chose which option you want (Anthropic API Key or alternatively your subscription) when you hit the subscription it will redirect you to a web page to connect to Claude (automatic)

@chretiensunited @TommyGriffith @AntoineRSX @openclaw How ?

Since I installed QMD, I never ran out of token on Claude! I don’t know what it does but my @openclaw loves it: github.com/tobi/qmd

@CryptoTekniker @openclaw just run this on your terminal and it will do the migration on its own, it is configured to do so : curl -fsSL openclaw.bot/install.sh | bash

@openclaw I am still using clawdbot from the original setup and all the commands are clawdbot … like this. Do I need my bot to update to openclaw ? Like do I give the link to openclaw github and say change your shell lobster

🦞 OpenClaw 2026.1.30 🐚 Shell completion 🆓 Kimi K2.5 + Kimi Coding: run your claw for free 🔐 MiniMax OAuth: one more model just a login away 📱 Telegram got a glow-up — 6 fixes from threading to HTML rendering Plus a bunch of community-contributed fixes across LINE, BlueBubbles, routing, security & OAuth. The lobster provides 😏 github.com/openclaw/openc…