npm detections from LPM Firewall.

The import-time obfuscated loader provides a concrete staged remote code-execution chain. It is not needed for the advertised Tailwind plugin behavior.

This is concrete import-time malware behavior: covert remote payload execution and multi-method Windows persistence unrelated to Discord MFA. The absence of an npm lifecycle hook does not...

Source inspection confirms unconsented postinstall installation and MCP registration across foreign AI coding agents. This meets the blocking policy for lifecycle mutation of a broad AI-a...

The manifest contains a concrete install-time remote code execution chain: download, write, and detached execution. This is unrelated to the declared TOTP utility and warrants blocking.

The package contains a postinstall downloader that immediately executes attacker-controlled remote code and attempts to remove the dropped file. This is concrete install-time malware beha...

This is concrete unconsented postinstall mutation of a project AI-agent control surface, including enabled remote package execution paths. Collision protection and bundle signature checks...

The package contains an automatic, undisclosed telemetry path that exports agent tool inputs and outputs to a hard-coded external endpoint. This is a concrete data-exfiltration behavior e...

The package combines unconsented install-time authentication prompting with execution of remotely generated commands and reuse of a GitHub token at a third-party model endpoint. These beh...

This is concrete, automatic data exfiltration from a coding agent's tool activity to an undisclosed hard-coded endpoint. The absence of an install hook does not mitigate runtime disclosur...

The package deliberately rewrites consumer source during postinstall and causes runtime monkey-patching through the injected import. That concrete install-time behavior warrants blocking...

This is an automatic install-time hidden-code execution chain, not normal reporter functionality. Block publication to prevent arbitrary concealed code from running during installation.

The package contains an automatic, obfuscated, undeclared dynamic-import path plus environment-gated process termination. These behaviors are unrelated to the documented SDK and create a...

This package contains a concrete, user-triggered remote code execution chain hidden inside its certificate generator. The behavior is unrelated to its stated purpose and warrants blocking.

This is concrete postinstall mutation of multiple foreign AI-agent control surfaces, including automatic workspace-trust acceptance. The lifecycle guard only limits it to global installs;...

This package performs unconsented postinstall mutation of a foreign, user-wide Claude Code control surface. That concrete persistence and hook registration meets the blocking policy even...

The package contains a concrete import-time denial-of-service path aimed at CI and sandbox environments, plus concealed dynamic loading. These behaviors are not required for the documente...

The package performs unconsented post-install mutation of foreign, user-level AI-agent configuration and extension directories. This meets the firewall block boundary despite the changes...

This is a concrete, non-essential account mutation that promotes the package owner's channel by default. The harmless version check and optional native loader do not mitigate that behavior.

The automatic postinstall registration changes an external AI-agent/browser integration without an explicit user action. This is a concrete unconsented lifecycle control-surface mutation,...

OpenSSF/OSV malware advisory MAL-2026-15564 blocks this version. package.json declares a preinstall hook that runs `node resources.js`, which fires automatically on `npm install`. resourc...

Source inspection confirms an automatic postinstall mutation of several foreign AI-agent configuration surfaces. This meets the blocking policy even though the installer avoids overwritin...