The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs system discovery and security auditing by executing standard OS commands such as uname, ss, lsof, and firewall management tools (ufw, pfctl). These actions are strictly gated by a requirement for explicit user permission before execution.
[COMMAND_EXECUTION]: Facilitates the creation of scheduled security tasks using the openclaw cron utility. The skill design ensures that no persistence mechanisms or scheduled jobs are established without explicit user approval.
[REMOTE_CODE_EXECUTION]: Checks for software updates by querying the NPM registry using the npm view command. This is a standard administrative function for maintaining software version status and does not involve executing untrusted remote code.
[DATA_EXFILTRATION]: Although the skill reviews system configurations, it includes robust instructions to redact all sensitive information—including tokens, hostnames, and credentials—from its outputs and internal memory logs.

healthcheck