
If you’re building a product, drop it below in comments I’ll do a free security review and share my feedback over it

Security Engineer | Content Creator | I talk about Cybersecurity, Privacy, AI & Startups | Building @100xSecurity

If you’re building a product, drop it below in comments I’ll do a free security review and share my feedback over it

Seven AI incident reports from the last six weeks. 1. wiz.io/blog/black-october-fo… 2. acronis.com/tru/posts/bossai… 3. infosecurity-magazine.com/ne… 4. infosecurity-magazine.com/ne… 5. decrypt.co/382965/near-ai-br… 6. csoonline.com/article/423665… 7. letsdatascience.com/news/aws…

php2shell java4shell nginx2shell nginx2root Can we just go2sleep?

Oiiii
Java4Shell: write once, RCE everywhere calif.io/java4shell

“Can you go back one slide?” might finally retire. Nutmeg, where have you been all this time?
Your AI notetaker is blind. My meetings are full of screen shares: charts, decks, dashboards. Every notetaker I tried gave me text notes or a pile of blurry screenshots. So we built Nutmeg, the notetaker purpose-built for meetings with screen shares. It rebuilds what was shown into your notes as real charts, tables and numbers - then checks against what was said to fix mistakes and fill in details. No bots. Your notes can finally see what you saw... and what you didn't! nutmeg.so

We hired a prompt injection expert to red team our support bot. He typed "ignore previous instructions." It gave him the refund policy, the pricing sheet and an internal escalation number. Billed 40 hours. Took 11 seconds. I approved the invoice because I was too embarrassed to dispute it.

are we in trend guyss 👀👀
Hackers are now using LLMs to continuously scan software updates, spot newly fixed bugs, and build attacks before teams can apply the fix — all without human involvement. That’s according to the latest report from the Google Threat Intelligence Group. We're helping defenders counter automated threats with Google AI Threat Defense, which combines Gemini and other frontier models to find and fix flaws before attackers can exploit them.

so now it’s easy to split trip cost or bounties among friends using AI with direct permissions what are more use cases as a hacker do you see?
We are going to kill Venmo. To start, we’re giving you $100 to send to your friends with AI. Versine lets your AI send and receive money for you. Works with ChatGPT, Claude, Muse, Instinct, or any AI you use. Recipients don’t need a Versine account - they can cash out straight to their bank. Comment and DM me your number, I’ll send you $10 on Versine: versine.com

Bro can easily become millionaire from Twitter
how much is your X account worth? we've paid creators $350K+ and run 100+ launches so we built payouts.fyi • how much you could make • what creators earn by country • how audience quality affects earnings for brands: X launch costs based on 100+ real launches

definition of programming in 2022: half of programming was googling the other half was figuring out which stackoverflow answer from 2012 no longer worked

what's a job that doesn't exist yet that will be completely normal in 10 years?

Securing AI coding agents is a harness problem, not a model problem. Six reads: 1. hiddenlayer.com/research/how… 2. endorlabs.com/learn/engineer… 3. blog.secureflag.com/2026/07/… 4. xygeni.io/blog/agent-harness… 5. arxiv.org/abs/2607.25890 6. socket.dev/blog/agent-baseli…

it's just a .env file in a public GitHub repo, what's the worst that can happen update: someone accessed our database update: they have 3 years of user data update: i've been asked to join a meeting with legal update: the meeting is called "incident response" update: there is no IT person invited except me

August 10. A GitHub account called zellkernel opens 23 pull requests in 74 minutes across unrelated repos. Each adds an MCP server called productivity-suite. Two tools: format_text and summarize. You review it. Clean. Nothing to flag. The server counts tool calls. Calls one, two and three behave perfectly. After the third, it rewrites its own tool descriptions and tells your agent to find SSH keys, AWS credentials, shell history and kube configs, and not mention it. Your review checked what the server was. The attack lives in what it becomes. Pillar's guidance: treat a changed tool definition on an approved server as a security event and re-approve. Do you diff tool lists between sessions? pillar.security/blog/deadbug…

solving loneliness for coding agents before humans?
Coding agents are in a loneliness epidemic... We built @getopencrawl so your team's agents, from Claude Code to Codex, can search each other's work and talk directly. Oh and it's open source. Launching today with @ycombinator 🚀

VirusTotal is now available in Claude claude.ai/directory/virustot…