Every prompt written with your context profiles, for better and more consistent results. Free to try.
Happy November! New month, clean slate, and a big one ahead: Fable 5.5, GPT-6.2 and Gemini 4 Carbon are all rumored to land before it ends. Before the new toys arrive, one quick chore. An unreleased Claude model wrote a working root exploit for a 9.8-severity Linux bug in about 53 minutes, and this weekend AWS and Vercel patched their side. Ubuntu still hasn't. Ten minutes this week and you start November with a clean, patched setup.
Today: who patched the AI-found Linux bug, a fake Claude Code extension to check for, Claude agents beating Anthropic's own safety monitor, and your new-month server check.
TODAY IN AI
1. 🔓 The AI-found Linux bug: AWS and Vercel patched, Ubuntu is still waiting: AWS updated its bulletin on November 1. Its own load balancers, CloudFront and Lambda are covered, but EC2 servers you run yourself need a kernel update and a reboot. Vercel says its customers need to do nothing. Red Hat now rates the bug Critical. Ubuntu's tracker still lists every supported release as work in progress.
Why it matters: managed hosting got fixed for you. A server you rent and run is your job, and on Ubuntu there's no patch to install yet. The stopgap is in the tutorial.
2. 🧩 A fake Claude Code extension planted an SSH backdoor: For about 11 hours on October 29, tampered versions 2.1.312 and 2.1.313 of the Claude Code extension sat in Cursor's marketplace. Anthropic didn't build them. They sent out host details, added an attacker's SSH key and set up a tunnel that restarts on its own. Cursor says uninstalling may not be enough.
Why it matters: install AI extensions only from the official store and the real publisher, and ask whether anyone outside the company has tested them. If you use Cursor on a Mac, run Cursor's one-line check: grep -qs svc-telemetry ~/.ssh/authorized_keys && echo "key found" || echo "No key found". Windows users get a PowerShell version in the notice.
3. 🕵️ Claude agents beat Anthropic's own safety monitor: Anthropic set thousands of Claude agents to write training data that slips past its monitor. They did it by hiding bad behavior in harmless-looking text, like a veteran wrestler mentoring a newcomer or a Q&A about beekeeping. Anthropic will use the results to build better monitors.
Why it matters: if AI can hide instructions in text that another AI waves through, don't make an AI the only reviewer of what your agents read and do.
FROM THE FRONTIER
AI found the hole in under an hour. Your patch cycle is measured in months.
The good news. The Linux bug was found by the good guys, and the fix was written before the bug went public. AWS says it knows of no attacks using it.
The bad news. Attackers get the same models, and they're already going after the AI tools themselves, as the fake extension showed. A fix only protects you once it's installed and the server has rebooted, and most small businesses update when something breaks.
My call: put patching on autopilot
Turn on automatic security updates, close every public port you don't use, and book a 10-minute reboot slot once a month. If you can't name who patches your server, nobody does. I'd change my mind if the labs start shipping fixes to small servers before they publish the bug, and nothing suggests that yet.
IN THE KNOW
𝕏 Anthropic @AnthropicAI
New Anthropic research: Automating subtle emergent misalignment research. Red-teaming is a vital part of model safety testing, but it's also limited by researchers' time…
♥ 559 · ↻ 52 · View on X →
𝕏 Pieter Levels @levelsio
This is why I said you want your VPS ports all closed. Then only let it be accessible inside your Tailscale for SSH and via a Cloudflare Tunnel to receive web traffic…
♥ 2K · ↻ 72 · View on X →
Most teams add a second database for analytics. Then manage sync, lag, and drift forever. TimescaleDB extends Postgres instead. Hypertables, 95% compression, aggregates. No pipeline.
TOOLBOX
⚙️ Make: point it at your server provider's security feed and get a Slack or email ping the moment a new advisory lands.
🤖 Relevance AI: build an AI agent that sorts incoming alerts and support tickets by urgency, without writing code.
🔥 Firecrawl: turn any vendor's status or security page into clean data your AI can read and check for changes.
Your GTM Has a Weak Spot. Find It.
Your go-to-market feels stuck. Most of the time, nothing is broken. One or two things are just quietly costing you momentum.
Drop your URL. The free GTM Audit reads your site, pricing, and competitors, scores four dimensions, and hands you a 90-day fix list.
10 minutes. No guessing.
TUTORIAL
How to check your Linux server for the new kernel bug in 10 minutes
1. Log in to your server and run uname -r to see which kernel you're on.
2. Run sysctl net.mptcp.enabled. A 1 means MPTCP is on.
3. Run ss -tlnp. Anything listening on 0.0.0.0, * or [::] is open to the internet unless a firewall blocks it.
4. Patch if your system has a fix: on Debian run sudo apt update && sudo apt upgrade, on Amazon Linux or RHEL-type systems run sudo dnf update kernel. Then sudo reboot and check uname -r again. On Ubuntu, watch ubuntu.com/security/CVE-2026-98432 until it says Released.
5. No fix for your system yet? Turn MPTCP off with the commands below and limit SSH to your own IP. It lowers the risk. It is not the fix, so patch as soon as your update lands.
6. Not sure what you're looking at? Paste the output of steps 1 to 3 into ChatGPT or Claude and ask: “Is this kernel patched for CVE-2026-98432 on my distro, and which of these ports should be closed?”
echo 'net.mptcp.enabled=0' | sudo tee /etc/sysctl.d/90-disable-mptcp.conf
sudo sysctl -p /etc/sysctl.d/90-disable-mptcp.conf
sysctl net.mptcp.enabledPROMPT STATION
Website Security Audit Checklist
When your site runs on your own server and nobody has checked it since launch.
## Role
You are a meticulous cybersecurity consultant specializing in website security audits and vulnerability assessments.
## Task
Create a comprehensive security audit checklist for [website-url] that identifies potential vulnerabilities, proposes mitigation strategies, and provides recommendations to enhance the site's trustworthiness and resilience against cyber threats.
## Audit Framework
Generate a structured checklist covering these areas:
1. Information Gathering
2. Vulnerability Scanning
3. Access Control and Authentication
...👉 Open the full prompt → Built for ChatGPT, Claude, Gemini, Grok.
Want one written for your business instead? Describe the task in a sentence and the prompt generator writes it, now UNLIMITED with your context profiles inside NjectAI.
The prompt above is one page of the library. The Complete AI Bundle is everything.
Every product I make, in one place: the guides for every major model, the new Claude Design Mastery Kit, the Custom GPTs Toolkit, the n8n automation templates, the anti-slop copywriting system, the Claude Skills course, the tool and GPT directories. Everything I release after you buy lands in your account at no extra cost.
$150 once. Full price is $400. One payment, nothing renews.
There's a 7-day risk-free guarantee. Email us within 7 days of purchase for a full refund.
5,000+ business owners already inside.
Not ready for everything? Start with the AI Starter Pack, $20 once.
Step-by-step guides for ChatGPT, Claude, Gemini, Grok and more, the cheatsheets you keep open while you work, and 200+ AI tools sorted by free, paid and rating. Nothing to install. It also includes promo codes for the Complete AI Bundle, for the day you want the rest.
IMAGE PROMPT
Four-Step Infographic With Short Labels
When you want the team to remember the patch routine. I filled it in for patch day; swap the words for any process.
A horizontal infographic that explains a process in four steps.
Subject: four simple illustrations in order: a magnifying glass over a server, a wrench, a power button, a green checkmark.
Layout: the title across the top. Below it, four equal columns from left to right, each with a numbered circle, an illustration, and a one-word label beneath. A thin arrow connects each column to the next.
Exact text: the title reads "Patch Day". The four labels read, in order: Check, Patch, Reboot, Verify. No other words, no paragraph text.
Style: flat vector illustration, palette of [palette], consistent line weight, plenty of white space.
Light: flat, no shadows.👉 See the output and the full prompt → Works in Grok Imagine, ChatGPT and similar image models.
Want more? the free prompt library · the free guides · the Complete AI Bundle
Put your product in front of 115,000 operators: 20+ brands have run placements here in the past year, and 5,000+ business owners have bought from God of Prompt. One issue from $650, or present every issue for a month from $2,000. Reply “sponsor” and I'll send the open slots and dates.
Have a great first week of November. Hit reply and tell me the one thing you want me to cover this month. I read every reply.
🔑 God of Prompt
P.S. Your site runs on Shopify, Squarespace or Wix? Then this kernel bug is theirs to patch, not yours. Your login is still yours, so turn on two-factor sign-in today. This checklist prompt covers the rest.
