Urgent.News

What's breaking now, across thousands of outlets.

Tech

Worm armed 42 GitHub repos by hiding a payload too big to fail

Payload blobs measuring 12.9 to 13.7 MB were large enough to evade GitHub code search

Worm armed 42 GitHub repos by hiding a payload too big to fail

A new Miasma-family npm worm, initially believed to be a harmless test without any victims, has secretly infected 42 GitHub repositories from 11 owners. The stealthy payload was designed to evade detection by GitHub and automated scanners, as reported by JFrog researchers. The threat was first identified by SafeDep, an open-source supply chain security firm, which discovered 18 packages associated with the operator between October 19 and October 25.

The malware self-replicates, steals credentials, and installs a wiper that deletes the victim's home directory when they revoke the stolen GitHub token. The new worm shares code with a June Miasma Phantom Gyp worm and adds Mini Shai-Hulud's token-revocation feature. However, SafeDep found no compromised third-party packages or victims as of October 22.

JFrog's analysis revealed that the operator had already armed real repositories belonging to established developers and organizations. The attackers embedded execution hooks for AI coding assistants and development tools, like Claude, Gemini, Cursor, and VS Code, alongside polymorphic wrappers ranging from 12.9 to 13.7 MB that decode hidden CI-publisher payloads.

They also stole credentials, modified repositories and CI configurations, and spread through npm packages. To avoid detection, the malware checks for CI runners, containers, sandboxes, debuggers, and other signs of automated analysis before launching a loader that pulls a second payload from the operator's own GitHub account. This account, CementBeetleSteeple, hosted 17 campaign repositories in total and is connected to a shellpanel scanner previously documented by Socket.

Written by urgent.news from The Register's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at theregister.com →

More in Tech

Debug Log #10: A Post Didn't Publish On Schedule, and I Can't Fully Explain Why

One post out of thirty didn't go out on schedule, and the exit code it left behind means "something killed this process," not "something went wrong inside it." I found this by accident, checking in on…

  • One scheduled post failed to publish on time, 36 hours late
  • Machine was asleep or offline at designated publishing time
  • Windows Task Scheduler log disabled, cause of termination unclear

What Did They Have to Check or Redo?

An AI-drafted update, summary or email can cost the person who gets it time you do not see. Ask them one question, then change one thing. Your AI drafted the weekly update in three minutes.

  • Recipients spend nearly two hours per incident fixing AI-generated update issues.
  • 50% view the sender as less capable after workslop.
  • Asking recipients to report what they had to check or redo can improve the process.

More from Monday 26 October →