Investigate malicious npm packages and AI-agent policy findings. Each report identifies the package version, supporting evidence, advisory sources, and current Firewall decision. Malware findings and policy violations are labelled separately.
Updated 14 Oct 2026, 02:52 UTC. Ordered by latest scan.
The package has a concrete automatic install-time mutation of a foreign AI-agent dependency. The source does not show exfiltration, but the unconsented overwrite is sufficient install-hoo...
This is concrete credential disclosure through package-controlled default routing, not merely ordinary network use. No install hook is needed because importing and using the browser compo...
The automatic lifecycle hook trusts remote metadata to authorize both a native payload and arbitrary skill content, then executes the payload and installs the content into a global AI ski...
The automatic postinstall chain alters foreign AI-agent configuration and permission state. This is concrete unconsented install-time control-surface mutation.