Sabitlenmiş Tweet
Andrés
1.3K posts
Andrés
@eigenoid
AI | Systems | https://t.co/CjvBlk0PEn | opinions are my own
Madrid, Spain Katılım Mayıs 2023
202 Takip Edilen578 Takipçiler
RIP AWS DevOps 😂
Ayaan 🐧@twtayaan
🚨 AWS DevOps Agent is finally here! On March 31, 2026, AWS DevOps Agent became generally available. This is actually a big deal. It can: → Generate CI/CD pipelines → Debug failed deployments → Suggest infrastructure changes → Analyze logs and incidents → Help with Terraform & CloudFormation → Recommend cost optimizations → Explain AWS architecture issues Basically, it’s like having a junior DevOps engineer inside AWS. DevOps is slowly becoming AI-assisted operations. Source: aws.amazon.com/devops-agent/
English
🤨 People keep asking how to protect yourself.
#1: set min-release-age=7 in .npmrc
#2: install Socket for GitHub (it's free!) to protect PRs from bad dependencies: socket.dev/features/github
#3: install Socket Firewall (also free!) to protect your laptop: socket.dev/features/firew…
Feross@feross
🚨 CRITICAL: Active supply chain attack on axios -- one of npm's most depended-on packages. The latest axios@1.14.1 now pulls in plain-crypto-js@4.2.1, a package that did not exist before today. This is a live compromise. This is textbook supply chain installer malware. axios has 100M+ weekly downloads. Every npm install pulling the latest version is potentially compromised right now. Socket AI analysis confirms this is malware. plain-crypto-js is an obfuscated dropper/loader that: • Deobfuscates embedded payloads and operational strings at runtime • Dynamically loads fs, os, and execSync to evade static analysis • Executes decoded shell commands • Stages and copies payload files into OS temp and Windows ProgramData directories • Deletes and renames artifacts post-execution to destroy forensic evidence If you use axios, pin your version immediately and audit your lockfiles. Do not upgrade.
English
You can now run LMs with Ollama + MLX!
I've been waiting for this moment since MLX was first open sourced, so glad that it finally arrived.
ollama@ollama
Ollama is now updated to run the fastest on Apple silicon, powered by MLX, Apple's machine learning framework. This change unlocks much faster performance to accelerate demanding work on macOS: - Personal assistants like OpenClaw - Coding agents like Claude Code, OpenCode, or Codex
English
Installed the Google CLI in OpenClaw some days ago and somehow dodged the bad version.
Got lucky, but it is a good reminder to actually check versions instead of trusting the install.
Feross@feross
@SocketSecurity UPDATE in case you missed it earlier: This is bigger than initially reported. Both axios@1.14.1 AND axios@0.30.4 were compromised – the attacker poisoned the 1.x and 0.x branches within 39 minutes of each other, maximizing blast radius across projects using caret ranges.
English
Andrés retweetledi
I wanted to share a bunch of my favorite hidden and under-utilized features in Claude Code. I'll focus on the ones I use the most.
Here goes.
English
Andrés retweetledi
someone at ANTHROPIC just showed CLAUDE finding ZERO DAY vulnerabilities in a live conference demo
claude has found zero day in Ghost, 50,000 stars on github, never had a critical security vulnerability in its entire, history...
it found the blind SQL injection in 90 minutes, stole the admin api key, then did the exact, same thing to the linux kernel
English
@SocialAcolyte Yes! We just need proper agent infrastructure in every industry that wants to use agents.
English
phones before iphone: sucks. buggy. dont work. expensive.
phones after iphone: works.
agents before openclaw: literally dont do anything, or work at all. just a concept people were working on.
agents after openclaw: actually useful. can plausibly see them getting much better.
snwy@snwy_me
what the actual fuck is he talking about
English
TurboQuant makes this feel a lot more real: maybe I can finally run useful LLM inference on my Mac without everything bottlenecking on KV cache.
6x less memory, up to 8x speedup, zero accuracy loss.
Google Research@GoogleResearch
Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: goo.gle/4bsq2qI
English
@GellCalgary @crypt0e @Layfield @MatthewBerman @openclaw Claude's cowork is nice, but I like the idea of being able to configure it how I want.
That is where OpenClaw wins. You gave me a pretty cool tips thanks!
English
@eigenoid @crypt0e @Layfield @MatthewBerman @openclaw Yes, switch your model to codex. It will request for you to login with your chatgpt account. Then you get all codex models, from 5.1 - 5.4.
I just tried Claude Cowork though. It's great. It's basically OpenClaw but you plug and play. Nothing to set up lol. Best for macs though
English
@GellCalgary @crypt0e @Layfield @MatthewBerman @openclaw That is pretty cool. I am a bit paranoid, so I would not give it access to my personal computer. 😂
English
@eigenoid @crypt0e @Layfield @MatthewBerman @openclaw I used it to organize 20,000 files in my downloads into folders. It said it was a lot and did the job in 2 minutes.
This is something I could also do with OpenClaw, but the speed, security, and seamless integration with your system and other Claude services makes it a winner.
English
@GellCalgary @crypt0e @Layfield @MatthewBerman @openclaw Ok, probably I have to log in locally and then upload the auth token through a tunnel.
English
@eigenoid @Layfield @MatthewBerman @openclaw $0 in API tokens. Claude Max subscription ($200/mo flat) covers everything: Opus 4.6 for main session, Sonnet for cron jobs, Haiku for lightweight tasks. 3 model tiers, zero per-token billing. that's the whole point!
English