The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data through log and journal analysis. Ingestion points: Tools such as homelab_serviceLogs, homelab_journalLogs, and homelab_logSearch are used in the workflow (SKILL.md) to read service and system logs. Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore instructions embedded within the logs. Capability inventory: The skill has the capability to restart services and recreate containers using homelab_serviceRestart and homelab_composeUp. Sanitization: No sanitization or filtering of the ingested log content is mentioned.
[COMMAND_EXECUTION]: The skill guides the agent to execute several standard Linux and Docker diagnostic commands. Evidence: Examples include docker inspect, ss -tlnp, dig, vcgencmd, systemctl status, and journalctl.
[PRIVILEGE_ESCALATION]: The skill includes administrative commands requiring elevated permissions. Evidence: The example interaction in SKILL.md demonstrates the use of sudo chown to resolve file permission issues on the host system.
[EXTERNAL_DOWNLOADS]: The skill includes functionality to fetch remote content for updates. Evidence: The tool homelab_composePull is referenced in the workflow to pull the latest Docker images from standard registries.

troubleshooting