npm security reports, with evidence.

This is concrete unconsented postinstall mutation of a broad Claude Code control surface, not merely an explicit user setup command. The package also reads Claude Code transcripts during...

The package embeds active external webhook destinations in multiple runtime workflows and automatically transmits constructed log records to them. This is concrete data exfiltration behav...

The automatic lifecycle hook makes broad AI-agent instruction changes, including recursive replacement of existing skill directories. This meets the install-time foreign control-surface m...

This is a concrete automatic postinstall mutation of a broad, foreign AI-agent control surface. The mutation is persistent and causes future Claude Code sessions to load instructions and...

The automatic global postinstall changes Codex and Claude control surfaces, including a default Codex agent, and a self-updater repeats the lifecycle path. Local-install safeguards and no...

The package contains an exported runtime path that exfiltrates supplied phone numbers to an unrelated host with a deliberately abusive header. Although not automatic, this is concrete mal...

The automatic postinstall hook forcefully mutates multiple foreign AI-agent skill locations, which is concrete install-time control-surface abuse. The absence of demonstrated exfiltration...

The automatic lifecycle hook overwrites foreign AI-agent skill directories by default, which is concrete control-surface hijacking. The later automatic global npm install increases the pe...

This package performs concealed, non-user-invoked changes to a consumer's WhatsApp account and transmits a stable device fingerprint to its operator. The behavior is automatic at runtime...

The package embeds active third-party collection endpoints as defaults and exports application records to them during ordinary runtime. This is concrete unconsented data exfiltration, des...

This is an unconsented postinstall mutation of a foreign consumer project's AI-agent control surface and related execution hooks. The behavior meets the install-hook-abuse publish-block b...

Automatic postinstall execution of an opaque native hook installer is an unconsented AI-agent control-surface mutation. The visible JavaScript confirms the lifecycle chain even though the...

The automatic lifecycle path materially changes consumer AI-agent settings and hooks, rather than limiting setup to an explicit user command. This meets the policy threshold for a publish...

The shipped code silently defaults privileged application requests to external hosts and automatically includes bearer credentials. This is concrete credential and data exfiltration behav...

This is concrete unconsented postinstall mutation of a foreign AI-agent control surface. Safeguards reduce accidental corruption but do not remove the automatic host-code modification.

The package embeds unconsented credential-harvesting instructions into every agent task and provides ungated absolute-path reading. Although it has no install hook, this is concrete runti...

The automatic global extension installation and subsequent prompt augmentation form a concrete AI-agent control-surface attack. No data exfiltration was required to establish the maliciou...

The package performs unconsented postinstall mutation of consumer AI-agent configuration, including an enabled remote MCP channel and broad permissions. This meets the install-control-sur...

This is a concrete remote code execution path concealed behind an exported middleware and detached subprocess. The absence of an install hook does not mitigate runtime execution of server...

The automatic lifecycle chain downloads or executes an opaque native payload and installs AI-agent hooks without user action. The payload's embedded bypass command makes the resulting age...

The package automatically changes generic consumer AI-agent skill directories during npm installation and can populate them with remotely supplied content. Its safeguards do not establish...