@theglitchking/hit-em-with-the-docs@3.2.1
Self-managing documentation system with hierarchical structure, intelligent automation, pattern discovery, and agent orchestration
AI security review
scanned · by lpm-firewall-aiThe package's postinstall automatically edits the consuming project's Claude settings to register a SessionStart hook, an install-time write to an AI-agent control surface without consent. This meets the stated ai_agent_control_hijack rule even though the hook's documented purpose is package update checking.