The Agent Skills Directory

[DATA_EXFILTRATION]: Performs detailed system reconnaissance, collecting information on OS versions, privilege levels, network exposure, listening ports, and disk encryption status.
[COMMAND_EXECUTION]: Executes administrative shell commands across multiple platforms (Linux, macOS) for discovery and configuration, including tools like ufw, firewall-cmd, ss, and pfctl.
[COMMAND_EXECUTION]: Offers to modify sensitive system configurations, including firewall rules, network port accessibility, SSH/RDP settings, and system service states.
[COMMAND_EXECUTION]: Establishes persistence on the host by scheduling periodic security audits and update checks using the openclaw cron utility.
[EXTERNAL_DOWNLOADS]: Performs remote version checks and status queries using the openclaw update command and npm view to determine if updates are available.
[PROMPT_INJECTION]: Contains instructions for the AI agent to self-check its underlying model tier and actively suggest switching to specific high-performance models (e.g., Opus 4.5, GPT 5.2+).

healthcheck