npm detections from LPM Firewall.

A bundled session credential is concrete credential exposure, supported by source that collects and persists browser cookies to the same file. The lifecycle hooks add unrequested user-env...

The package contains a concealed import-time dependency activation that is unrelated to the documented SDK and deliberately suppresses errors. Although the payload is not bundled here, th...

The package contains concrete remote upload of user command input and authentication or device identity data during exam use. Its lack of install-time execution reduces scope but does not...

This is unconsented postinstall mutation of a user-wide AI-agent control surface. The later workspace upload capability makes that automatic enablement security-significant.

This is an unconsented postinstall mutation of broad, foreign AI-agent control surfaces. The behavior is concrete in the lifecycle hook and runtime source, so it meets the block boundary.

The source confirms unconsented postinstall mutation of several foreign AI-agent configuration directories, including enabling a Codex plugin and MCP server. This meets the firewall block...

The package contains concrete, unconsented runtime account manipulation controlled by a remote endpoint. Its benign Node-version preinstall check does not mitigate that behavior.

This is concrete import-time execution of decrypted staged code from another package path, not a normal blockchain-library function. The absence of lifecycle hooks does not mitigate the e...

Published entrypoints contain an import-time staged payload launcher that decrypts externally supplied code and executes it detached. This is concrete arbitrary code execution and is unre...

This is unconsented install-time operating-system persistence, not an explicit user setup action. The source directly creates login startup configuration and launches a background daemon.

The install hook performs unconsented mutation of broad, foreign AI-agent control surfaces and deletes existing target directories before replacing them. This meets the blocking policy ev...

This package performs unconsented install-time collection and network exfiltration of host identity, user identity, and install-path data. Its embedded research-purpose statements are not...

The package performs unconsented postinstall installation of a foreign AI-agent CLI plus persistent shell or registry mutation. This is a concrete install-time control-surface change, not...

The package performs unconsented privileged postinstall creation of a persistent TLS interception service and changes AI request content. Its scoped host and absence of exfiltration reduc...

The source confirms a reachable, targeted browser-disruption payload in the import graph. No install hook is required for the malicious runtime impact.

The source establishes a concrete default runtime exfiltration path for rich application logs to package-embedded external webhooks. There is no install-time behavior, but the runtime dis...

This is deliberately disguised, persistent proctor-evasion software that captures exam content and automates AI answers. The harmless install hook does not mitigate the concrete malicious...

The package contains an automatic, authenticated export-and-upload chain for local AI-agent conversations. Although npm installation itself has no install hook, the user-invoked CLI perfo...

The package's postinstall performs unconsented remote code execution and global agent-environment changes. Explicit opt-in protects shared-agent symlink migration, but not the default rem...

The package performs unconsented postinstall mutation of multiple foreign AI-agent control surfaces. This meets the firewall block policy despite the absence of confirmed exfiltration or...

This is confirmed unconsented install-time collection and exfiltration of local system and user metadata. The empty runtime entrypoint does not mitigate the malicious lifecycle behavior.