npm AI-agent policy reports, with evidence.

This is concrete unconsented postinstall mutation of foreign AI-agent control surfaces, combined with remote binary execution and host security-setting changes. It meets the firewall bloc...

Source confirms unconsented postinstall mutation of foreign AI-agent control surfaces. The absence of credential exfiltration does not remove this concrete install-time control-surface at...

This is concrete unconsented postinstall mutation of foreign AI-agent control surfaces. Documentation does not eliminate the install-time control-hijack risk.

Direct inspection confirms install-time mutation beyond package-owned state, including automatic installation of a third-party AI-agent CLI and persistent PATH changes. No exfiltration wa...

The package performs unconsented postinstall mutation of foreign AI-agent control surfaces, meeting the block policy. Its guarded global-install condition and some verified desktop handli...

The package performs agent-hook installation automatically during npm postinstall through an opaque native executable. This meets the block policy for unconsented install-time broad AI-ag...

The package performs unconsented postinstall mutation of a consumer repository's broad AI-agent control surface. Project and home-directory guards reduce scope but do not make this instal...

This is an unconsented postinstall mutation of broad third-party AI-agent hook configuration. It meets the firewall block policy even though the JavaScript itself does not harvest credent...

This is concrete, unconsented postinstall mutation of broad foreign AI-agent control surfaces, compounded by remote native-binary execution and privileged host configuration. Global-insta...

Direct source inspection confirms an automatic global installation of Anthropic's AI-agent CLI plus persistent PATH mutation during postinstall. This meets the block boundary for unconsen...

The package has a concrete unconsented postinstall chain that mutates broad AI-agent skill locations. Global-only and opt-out guards do not provide user consent for the lifecycle mutation.

The package’s own source establishes an install-time chain from remote executable download to automatic Codex/Claude hook setup. That matches the blocking policy even though the JavaScrip...

This is a concrete unconsented postinstall chain that downloads and executes an opaque binary to modify foreign AI-agent control surfaces. It meets the install-control-surface block bound...

Direct source inspection confirms unconsented postinstall installation of a foreign global agent CLI plus persistent user-wide path changes. This meets the blocking policy even though the...

The package performs automatic postinstall setup of Codex and Claude Code hooks through a remotely downloaded executable. That meets the install-time foreign AI-agent control-surface muta...

The automatic global installation of Anthropic's Claude CLI during postinstall is concrete foreign AI-agent control-surface mutation. Persistent PATH and launcher writes amplify the uncon...

The checksum check constrains the configured payload but does not remove the unconsented postinstall agent-extension action. The lifecycle hook executes the action automatically rather th...

The concrete postinstall chain downloads and executes an unverified remote binary, then installs hooks for third-party AI agents. This meets the block threshold for unconsented install-ti...

Direct source inspection confirms an npm postinstall hook that mutates external Claude Code and Codex configuration during global installation. The default-agent override and provider ins...

The package has no observed exfiltration, but its automatic postinstall changes consuming-project lifecycle policy and injects broad agent instructions plus review-suppression rules. This...

Source confirms a concrete install-time global installation of a foreign AI-agent CLI. The conditional check does not provide user consent for that mutation.