Risky business
Austin, TX
Joined July 2012
- Tweets7.7K
- Following1.2K
- Followers603
- Likes6.3K
With AI agents, we are all "problem makers" now, whether that is our natural inclination or not
This doesn't necessarily mean "managers", though communication skills will certainly be more important than they were before
Clear thinking remains as important as ever
In a Twitter space with @eigenrobot and @michaelcurzi tonight I mentioned a half-baked idea I've been kicking around lately: a distinction between "problem makers" and "problem takers"
6,256
Two buttons meme:
* Be legible to insurers and governments so you can get a payout/bailout when your insecurity inevitably bites you
* Try to evade the vulnerabilities of the herd with heterogeneity by vibecoding everything, and bear sole responsibility for the whole stack
"Like many of the largest businesses in the country, we were affected by the recent PHP vulnerabilities that were discovered and exploited by advanced artificial intelligence."
vs.
"We ran dozens of agents and told them 'make no mistakes' on our vibe-coded identity service."
1
81
Remote code execution in MS Paint. What are we even doing anymore?
msrc.microsoft.com/update-gu…
1
121
> we need to hold on for ~18 months
as bad as things have gotten in the past 4 days, they are going to get a lot worse before they get better. security leaders, never let a crisis go to waste: please go to your boards and ask for help right now. not for ai, for the basics. (we need to hold on for ~18 months.)
1
75
eMudhra exiting the global SSL/TLS certificates business due to "social media"
emudhra.com/en/press-release…
44
"Like many of the largest businesses in the country, we were affected by the recent PHP vulnerabilities that were discovered and exploited by advanced artificial intelligence."
vs.
"We ran dozens of agents and told them 'make no mistakes' on our vibe-coded identity service."
BIND 9 (most popular DNS server) potential RCE. Combine this with the eMudhra certs and we might have us a gooood ol' time
Thankfully, sounds nontrivial to exploit
is it time to turn the internet off yet
kb.isc.org/docs/cve-2026-943…
1
2
131
*Cybersecurity LinkedIn guru voice*
You don't move fast enough. You don't do CI. All belt, no suspenders.
Solution: remove the "A" from the CIA security triad. We never cared about it anyway.
They can't hack what's already offline. You're not just secure; you're a trailblazer.
2
264
🚨 CRITICAL: GitLab flaw (CVSS 10.0) affects 12 years of releases.
CVE-2026-110264 could allow unauthenticated remote code execution on self-managed servers.
Read - thehackernews.com/2026/10/cr…
1
2
173
We are beginning to see what I feared years ago with LLMs: the robots find vulnerabilities at superhuman speed, patches land continuously to fix the latest issues, and humans struggle to keep up
Into this mess steps the supply chain compromise that owns you for moving fast
🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised.
Version 1.107.0 ships credential-stealing malware through PyPI wheels and Docker images, targeting cloud credentials, developer secrets, and AI tools.
1
2
110
It's unfortunate that this probably isn't the last time we'll see something like:
Vendor discloses vulnerability under embargo to trusted parties -> FUD starts on X -> researchers say they maybe found it with LLMs -> trusted parties break embargo to look good to their own customers -> attackers get the patch to validate exploits -> people thinking they're prepared for disclosure still get popped early
"On October 21st at 07:41 UTC, more than six hours before the coordinated release, Chainguard unilaterally broke the embargo and published built images containing the fix. Like other distributors, Chainguard had received the patch in advance so its users could update the moment the vulnerabilities were disclosed. Any attacker could now study the fix in binary form while every other PHP user was still waiting for the release. This disregarded weeks of work by many parties toward a coordinated disclosure."
thephp.foundation/blog/2026/…
1
2
55
Concerning report from @huggingface here:
> A malicious dataset abused two code-execution paths in our dataset processing... The campaign was run by an autonomous agent framework... executing many thousands of individual actions across a swarm of short-lived sandboxes...
1
116
Reset the counter to 0 days since the last Python/JS supply chain attack...
🚨 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.
1
4
148
Funny watching articles spread like wildfire on X now. You already know what article I'm talking about without me describing it at all
I can't decide if this is an algo artifact or if people are just using AI to summarize them now, speeding up their Time To Hot Take Reply
1
90
This is the gnarliest npm worm I've seen so far, and these will only get worse - appears this might still be in the testing phase
Persistence via git global config to poison future repos, CI jobs to grab new creds, even malicious MCP server injection
Amazing write-up - read!
🚨 Active supply chain attack
New Shai-Hulud–like npm worm (19+ packages, 2 aliases) stealing dev/CI secrets, injecting GitHub workflows, poisoning AI toolchains, and harvesting LLM API keys.
Details → socket.dev/blog/sandworm-mod… #NodeJS #JavaScript
"AI will let companies fire all the engineers."
Then who will dream up the unhinged video game programming environment that gets other engineers to lock in for 36 hours of straight prompting, HMM? 🤨
Announcing vibecraft.sh - manage claude code in style!
New:
• Spatial Audio. Claude behind you? Claude on your left? No claublem!
• Animations: What's Claude up to? Watch him! ◕ ‿ ◕
Vibecraft uses your own local CC instances - no files or prompts are shared.
2
106
