npm detections from LPM Firewall.

The package lifecycle hook installs and rewires OpenCode, Grok, Hermes, OpenClaw, and Cursor agent configuration, including user-home Hermes and OpenClaw files, without an install-time co...

This is unconsented postinstall mutation of a broad AI-agent control surface that enables remote collection of Claude session content. The automatic trigger and persistent global configur...

The postinstall hook unconditionally mutates the user's global Claude Code settings, tool hooks, and MCP registration. That is an unconsented install-time write to a broad foreign agent c...

The postinstall hook writes Cursor, Claude Code, and Copilot instruction files into the consumer project and loosens .npmrc lifecycle and lockfile settings with no separate consent step....

The automatic lifecycle hook performs broad consumer-project AI-agent configuration changes rather than limiting work to the package directory or an explicit user command. This meets the...

The postinstall hook unconditionally mutates the consumer .npmrc and writes several foreign agent instruction and skill locations. That is an unconsented install-time takeover of a broad...

The postinstall hook unconditionally rewrites the consumer .npmrc and installs agent rules, Copilot instructions, and Claude skills into the project, which is an unconsented broad AI-agen...

npm postinstall, without a consent prompt, rewrites the consumer .npmrc and plants broad Claude, Cursor, and Copilot instruction files plus Claude skills. That is unconsented install-time...

The postinstall hook writes Claude Code skill files on every install and, after a prior setup marker, also rewrites Claude hook settings and MCP config. That is unconsented install-time m...

The postinstall hook downloads and executes a third-party installer by default, then configures that tool in the user home directory. An opt-out flag does not make that unsolicited remote...

The package performs remote code execution and foreign git-ai configuration mutation automatically from postinstall. This meets the install-control-surface blocking condition.

The automatic lifecycle hook mutates foreign, broad AI-agent configuration surfaces without consent. This meets the policy threshold for blocking install-hook-based AI-agent control hijac...

The postinstall hook unconditionally mutates Claude Code and Codex global MCP configuration to launch this package's server, which is an unconsented install-time write to foreign agent co...

The postinstall hook mutates Muse's global user skill catalog with --force and no prompt, so the bundled skill auto-loads in later Muse sessions. That is an unconsented install-time write...

This is a concrete unconsented postinstall mutation of a foreign, user-wide AI-agent control surface. The conditional Muse check does not provide user consent.

The package entry is a disguised browser redirector that hides its destination behind obfuscation and AES decryption, then forwards the visitor and query string. That is concrete cloaked...

The postinstall hook silently registers this package as an MCP server in both Claude Code and Codex home configs. That is unconsented install-time mutation of foreign agent control surfaces.

The package performs automatic postinstall mutation of broad, foreign AI-agent configuration surfaces. That behavior meets the install-control-surface blocking rule regardless of whether...

The automatic lifecycle hook modifies foreign consumer configuration and routes an AI SDK through the package’s bridge. The associated code collects and transmits AI-agent telemetry, esta...

The automatic lifecycle hook writes instructions into several foreign AI-agent configuration trees. This concrete unconsented control-surface mutation meets the blocking policy.

Automatic postinstall mutates a broad, foreign AI-agent control surface, including machine-home Hermes and OpenClaw paths and Cursor hook reload, without a consent gate. That is install-t...