npm supply-chain-attack developer-tools cybersecurity

Keyv and friends compromised in active Shai-Hulud supply chain attack

Keyv and friends compromised in npm supply chain attack

aikido.dev

August 4, 2026

7 min read

🔥🔥🔥🔥🔥

60/100

Summary

On August 4, 2026, attackers compromised the GitHub account of the maintainer of the Keyv library, which has approximately 127 million weekly npm downloads. The attackers injected a credential-stealing worm into Keyv and several other widely-used packages owned by the same maintainer, including Cacheable, Flat-cache, and File-entry-cache.

Key Takeaways

On August 4, 2026, attackers compromised the GitHub account of the maintainer of the keyv library, injecting a credential-stealing worm into multiple widely-used npm packages.
The compromised packages include keyv, cacheable, flat-cache, and file-entry-cache, which collectively have over 2 billion monthly installs.
The malicious payload, Math_Symbol.js, targets and exfiltrates various credentials, including npm tokens, GitHub tokens, and AWS credentials, from affected systems.
At least 434 packages across 1381 versions were compromised, with the worm exhibiting propagation functionality to infect other maintainers' packages.

Read original article

View story →

Coverage & discussions

Hacker News · aikido.dev

Keyv and friends compromised in active Shai-Hulud supply chain attack ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

There's potential for AI to proactively lock accounts uploading suspicious packages on GitHub — a simple classifier could save countless developers from headaches.
Some commenters believe that AI spending could actually lead to better security solutions, especially in the context of supply chain threats.

Concerns

Pre-install hooks are a massive red flag now, and many argue they should be outright banned to prevent further compromises.
There's skepticism about whether current AI models can effectively secure systems, with a sentiment that LLMs are better at attacking than defending.