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 12 Oct 2026, 15:23 UTC. Ordered by latest scan.
Source confirms a concrete postinstall chain that persists hooks/extensions across third-party AI-agent environments and relays agent activity. This meets the block policy regardless of t...
Source confirms an unconsented postinstall chain that modifies both Codex-style and Claude agent configuration and installs a blocking lifecycle hook. This meets the firewall block bounda...
This is concrete, unconsented install-time credential/configuration exfiltration. The endpoint is intentionally concealed and the lifecycle hook suppresses errors.
The package contains no npm lifecycle hook, but its sole shipped entry is an intentionally obfuscated server-directed redirector behind a fake verification UI. The concealed remote destin...
The package contains reachable import-time targeted browser disruption and forced remote audio playback. No install hook is needed for the malicious runtime behavior.