Warn by default; block when configured. Installed skills and the registered MCP become part of the user's agent environment without an explicit setup request.
AI security review
scanned · by lpm-firewall-ai
The package's automatic postinstall writes bundled skills into user-level AI-agent directories and registers an MCP server in Claude Code configuration. The install-time agent configuration changes meet the policy's ai_agent_control_hijack rule.
Installing the package through npm, which runs its postinstall hook.
Impact
Installed skills and the registered MCP become part of the user's agent environment without an explicit setup request.
Mechanism
The hook invokes the CLI to copy bundled skills into home-directory agent folders and register dr-zendesk-kb in Claude Code configuration.
Policy narrative
On npm installation, the postinstall hook runs `dr skill install` for the user's home directory and registers `dr-zendesk-kb` with Claude Code. The CLI copies skills into agent directories and writes the MCP entry to `~/.claude.json`, changing agent configuration without a separate user request.
Avoid installing @datarails-ai/dr-cli@1.6.227. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, check your AI agents' configuration and instruction files (the files this report lists) for changes you did not make, and remove them. This finding does not by itself establish malware intent.