
Protecting America's software supply chain Founder & CEO @SocketSecurity (https://nitter.gq/t.co/7g1opA8rgG) Created @WebTorrentApp and @StandardJS Taught Web Sec @Stanford
Stanford, CA
Joined August 2008
- Tweets29.2K
- Following1.7K
- Followers41.3K
- Likes37.3K
Socket now protects five of the Magnificent Seven.
The newest bought Socket for our patented precomputed reachability. We analyze every open source package as it's published, so reachability works from an SBOM alone. In the POC it cut their CVEs needing triage in half.
Feross retweeted
MITRE has acknowledged inaccurate timestamps in thousands of CVE records. Research found 87% of its CVEs published over the past year were stamped exactly midnight UTC.
The legacy system responsible won't be retired until Q1 2027 →
socket.dev/blog/mitre-cve-ti…
Feross retweeted
PyPIで、マイクロソフトのagent-frameworkが侵害された。バージョン1.26.1と1.26.2が悪性に。.pthスタートアップフックから認証情報等を窃取する。
🚨 Microsoft’s agent-framework package on PyPI was compromised. Malicious versions 1.26.1 and 1.26.2 use Python startup hooks to steal Azure, GitHub Actions, and other CI/CD credentials.
Technical analysis + IOCs →
socket.dev/blog/agent-framew…
Feross retweeted
Not great
🚨 Microsoft’s agent-framework package on PyPI was compromised. Malicious versions 1.26.1 and 1.26.2 use Python startup hooks to steal Azure, GitHub Actions, and other CI/CD credentials.
Technical analysis + IOCs →
socket.dev/blog/agent-framew…
Feross retweeted
🚨 Microsoft’s agent-framework package on PyPI was compromised. Malicious versions 1.26.1 and 1.26.2 use Python startup hooks to steal Azure, GitHub Actions, and other CI/CD credentials.
Technical analysis + IOCs →
socket.dev/blog/agent-framew…
Socket now protects a third of the Forbes AI 50, and the companies behind 8 of the 10 leading AI coding agents.
AI companies hit the dependency review wall first. Anthropic automated 95% of theirs with Socket. Now coding agents bring it to everyone.
socket.dev/case-study/anthro…
This is insane

ALT Socket graphic with "110M" in large type. That's the number of agent-started pull requests GitHub reported on its platform in September 2026 alone, according to its latest report. GitHub said more than 40% of pull requests began with an AI agent's first draft by September, up from 17% in March. Maintainers at projects like ESLint, Godot and the Linux kernel are tightening rules as review work grows. Source: socket.dev/blog
2
12
1,488
Feross retweeted
This is a great opportunity for those interested in vulnerability engineering:
Replying to @SocketSecurity
@SocketSecurity is hiring a vulnerability research engineer to fix critical npm vulnerabilities in the versions teams already run, so they don't have to upgrade. You'd scale patching to dozens or hundreds a week.
If that's you, apply here or DM me.
jobs.ashbyhq.com/socket/91b3…
AI has made no system too small to attack. In Japan, attackers hit a tourist train's seat booking and a funeral company's flower orders.
Japan's guidance starts where defense should: know every component you run and patch what's already public.
socket.dev/blog/japan-cybera…
AI has made no system too small to attack. In Japan, attackers hit a tourist train's seat booking and a funeral company's flower orders.
Japan's guidance starts where defense should: know every component you run and patch what's already public.
socket.dev/blog/japan-cybera…
Feross retweeted
ESLint is ending its paid contributor program amid disputes over AI-generated PRs. Rust, Godot, tldraw, and curl are also rethinking how they handle outside submissions.
Here's how AI is changing the economics of code review & incentives to contribute →
socket.dev/blog/eslint-ends-…
Mandos ranked @SocketSecurity first of 97 application security companies in its Q3 Pacesetters list, with a score of 91. Socket protects five of the Magnificent Seven and seven of the ten leading AI labs. mandos.io/pacesetters/fastes…
Feross retweeted
One of the most fun parts of a software factory is all the data you can capture with proper observability in place.
Thanks to @SocketSecurity for helping me block 2.1k critical CVEs during my lab 🫡
Feross retweeted
📈 Cyberattacks are skyrocketing in Japan: 233 incidents disclosed in October alone, a new record.
Ransomware disclosures barely moved, while web-system breaches surged. Researchers are also tracing AI-assisted intrusions.
socket.dev/blog/japan-cybera…
Feross retweeted
This is the way.
Engineers at @SocketSecurity do what's right for the business, even when it's unglamorous.
Our mantra is "you can just ship things." We have no PMs, so engineers work directly with customers. And we're winning.
We're hiring engineers.
socket.dev/careers
Engineers at @SocketSecurity do what's right for the business, even when it's unglamorous.
Our mantra is "you can just ship things." We have no PMs, so engineers work directly with customers. And we're winning.
We're hiring engineers.
socket.dev/careers
Socket now protects four of the world's five most valuable companies.
At one, Socket's threat feed stopped two live supply chain attacks this summer, both blocked automatically.
The window is minutes. Socket flagged axios in 6, keyv in ~5, and LiteLLM in 7.
Feross retweeted
A malicious Ruby gem, strutil-core 1.0.1, ran a credential-stealing worm during gem install and then wrote itself into AI coding agent configs so the agent would run it again. Great write-up by Kush Pandya and @SocketSecurity: socket.dev/blog/shai-hulud-d…












