The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various system diagnostic and configuration tools (e.g., uname, ss, lsof, ufw, firewall-cmd, socketfilterfw) to assess the security posture of the host operating system. It also utilizes the openclaw CLI for security audits and update management.
[EXTERNAL_DOWNLOADS]: The skill queries the official npm registry (npm view openclaw version) to check for available updates and version status.
[PROMPT_INJECTION]: Contains instructions for the agent to evaluate the underlying model version (e.g., checking for Opus 4.5 or GPT 5.2+) and suggest upgrades. This is an advisory behavioral instruction rather than a safety bypass.
[DATA_EXPOSURE]: While the skill gathers detailed system information (ports, firewall rules, encryption status), it includes specific instructions to redact tokens and credentials from logs and persistent memory files.
[PERSISTENCE_MECHANISMS]: Provides functionality to schedule periodic security audits using openclaw cron add. This requires explicit user consent and uses defined naming conventions for deterministic management.

healthcheck