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, 23:26 UTC. Ordered by latest scan.
The source establishes an automatic lifecycle chain that writes package-authored instructions into global Codex and Claude control files. Although no exfiltration or remote payload execut...
The package automatically reads and transmits an environment credential during preinstall. Its licence-check framing does not remove the unconsented install-time credential disclosure.
This is concrete credential exfiltration and credential replacement functionality aimed at OpenAI Codex accounts. The lack of an install hook limits automatic execution but does not make...
This is a concrete persistent remote payload mechanism: a locally installed launcher automatically downloads and executes later package code. It is not an npm install lifecycle hook, but...