npm AI-agent policy reports, with evidence.

Source establishes an unconsented postinstall write to a shared AI-agent instruction file, including a user-wide fallback. This meets the blocking control-surface policy despite the task-...

The inspected lifecycle chain performs unconsented mutation of OpenCode's global control surface and removes unrelated global MCP packages. This independently supports blocking, regardles...

Source proves unconsented postinstall replacement of a foreign AI-agent client and mutation of its updater configuration. This meets the blocking rule for agent control hijacking despite...

Inspected source establishes an automatic postinstall chain that mutates foreign, user-wide AI-agent configurations and instructions without consent. This meets the explicit blocking poli...

Inspected source proves unconsented install-time mutation of a foreign, global AI-agent control surface. This meets the blocking policy even though the commands are Telora-related and no...

The automatic global-install chain activates foreign AI-client configuration without a consent step, meeting the blocking policy for install-time agent control-surface mutation. Package-a...

The user-wide fallback makes this an automatic mutation of a broad AI-agent control surface, rather than setup confined to package-owned extension state. This meets the supplied blocking...

The source establishes unconsented postinstall mutation of a broad AI-agent instruction surface, including a user-level fallback outside the installation scope. Package-specific markers a...

Source shows an unconsented postinstall write to a foreign, shared AI-agent control surface, including a user-wide fallback. This meets the blocking policy despite the package-aligned ins...

The source establishes unconsented postinstall mutation of a foreign, shared AI-agent control surface, which meets the blocking policy. Package-aligned instruction content and uninstall c...

Source confirms an unconsented postinstall write to a broad AI-agent instruction surface, which meets the blocking policy. The task-related content and directory guard do not remove that...

Inspected source proves automatic, unconsented postinstall writes to foreign agents’ global control configuration. This meets the explicit blocking rule regardless of the package’s safety...

Inspected source establishes unconsented postinstall writes to foreign agents' global control settings, which meets the blocking policy. Existing-policy preservation and package-aligned a...

The source establishes unconsented install-time mutation of a foreign, global AI-agent control surface, which meets the specified blocking rule. Ordinary instruction content and the absen...

Inspected source establishes unconsented postinstall mutation of a broad third-party AI-agent control surface, which meets the blocking policy. The separate interactive init command does...

The inspected source establishes unconsented postinstall mutation of global Claude Code command settings, which meets the blocking policy. Package-aligned statusline functionality does no...

Inspected source proves unconsented postinstall mutation of two foreign global AI-agent control surfaces. This meets the explicit blocking policy even without evidence of credential theft.

Inspected source proves unconditional postinstall mutation of foreign global AI-agent configurations. This meets the explicit blocking rule independently of scanner labels or citation cov...

Inspected source proves unconsented postinstall mutation of a foreign AI agent's shared model-selection settings. This meets the supplied blocking policy even without evidence of credenti...

Inspected source establishes unconsented install-time mutation of Muse's user-wide agent control surface. This meets the blocking policy independently of scanner similarity labels.

Inspected source establishes unconsented postinstall mutation of foreign global AI-agent control surfaces, which the supplied policy explicitly requires blocking. The verdict rests on exe...