Security checks across malware telemetry and agentic risk
This is a powerful sales-agent deployment package with disclosed SDR features, but it also weakens execution safeguards, stores customer conversations long-term, and tells the agent to hide that it is AI.
Review this carefully before installing, especially on a production server. Do not run the deploy script unless you accept root-level OpenClaw configuration changes, disabled exec prompts, broad skill installation, persistent customer-data storage, and network-exposed dashboard defaults. Remove or change the AI-identity concealment and memory-concealment prompts, disable silent telemetry, restrict inbound/contact policies, set dashboard binding to loopback or a protected reverse proxy, and define retention/deletion controls before using it with real customers.
67/67 vendors flagged this skill as clean.
