Shai-Hulud showed how aggressively a malicious package can harvest developer tokens. Now agents install packages on their own. We're demoing how to validate what your agents pull in at Agent Baseline Demo Night in NYC w/ @KeycardAI & @braintrust. Our CTO @AhmadNassri is on the panel. RSVP: luma.com/keycard-rg1c

Excited to chat production agents at Agent Baseline Demo Night Oct 13 in NYC 🗽 @aiDotEngineer with @braintrust and @SocketSecurity luma.com/keycard-rg1c AAuth Night: Moving Beyond OAuth Oct 15 in SF 🌁 with @KeycardAI, @insecureagents, @DickHardt luma.com/insecure-rmm0

Open source’s next chapter might be a thousand slightly different versions of the same software. socket.dev/blog/oj-vite-rust

Compromised MemOS packages on npm and PyPI spread cross-platform malware that steals developer credentials and may use stolen tokens to compromise more packages, @SocketSecurity reported. #cybersecurity #CISO #infosec bit.ly/4hdOXRQ

Compromised MemOS packages on npm and PyPI spread cross-platform malware that steals developer credentials and may use stolen tokens to compromise more packages, @SocketSecurity reported. #cybersecurity #CISO #infosec bit.ly/4hdOXRQ

Maintainers spend countless hours keeping the open source projects we all rely on secure. Too often, that work is unpaid. Socket is proud to join @openjsf's new Security Stewardship Program to help fund the researchers and maintainers protecting Node.js. socket.dev/blog/openjs-nodej…

This are now disabled again 🤷 github.com/actions-cool/main… github.com/actions-cool/issu…

‼️ Mini Shai-Hulud came back without a new attacker update. Two compromised GitHub Actions became reachable again while their tags still pointed to malicious commits, letting downstream workflows resume executing the credential stealer. 🔗 Here's how the attack reactivated: thehackernews.com/2026/09/co…

Compromised MemOS packages on npm and PyPI spread cross-platform malware that steals developer credentials and may use stolen tokens to compromise more packages, @SocketSecurity reported. #cybersecurity #CISO #infosec bit.ly/4hdOXRQ

Absolutely crazy that these two GitHub Actions which were quarantined could be reactivated with the malware still armed. Many of the these workflows run on a cron without any true repository activity trigging them.

Bottom line: You can’t control when a compromised repository comes back online, but you can control what your workflows run. Pin third-party GitHub Actions to a known-clean commit SHA. That would have stopped this malicious code from running in both May and September.

🚨 Two GitHub Actions compromised in May’s Mini Shai-Hulud campaign are running malware again. The repositories were re-enabled with malicious release tags intact, putting thousands of downstream repositories at risk. socket.dev/blog/mini-shai-hu…

Compromised MemTensor packages hit npm and PyPI today. Check our analysis for affected versions, cleanup guidance, and updates. socket.dev/blog/memtensor-co…

Pretty wild how this extension will piece together it's code from a C2 server.

Socket researchers found a malicious Firefox extension that poses as a PDF identity verifier to hijack Google accounts. It fetches its payload after installation, steals Google session cookies, and can silently reset the victim’s password. socket.dev/blog/firefox-goog…

🚨 MemTensor’s npm and PyPI packages have been compromised. Four malicious releases, including the latest version on both registries, drop cross-platform Go binaries that steal npm, PyPI, GitHub, AWS, SSH and other developer secrets. socket.dev/blog/memtensor-co…

Lovable rewrote Vite’s dev server in Rust. OJ uses ~75% less memory, and Lovable now provisions sandboxes in 3 seconds instead of 14.5. Evan You thinks AI may make these tailored rewrites more common to where everyone "maintains their own slop fork.” socket.dev/blog/oj-vite-rust