dns-management
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEPRIVILEGE_ESCALATIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill includes instructions for administrative system changes, such as disabling
systemd-resolvedvia thesystemctlcommand to resolve port conflicts on the host machine. - [DATA_EXFILTRATION]: Instructions facilitate access to the AdGuard Home configuration file located at
/opt/homelab/docker/network/adguard/confdir/AdGuardHome.yamlfor managing DNS rewrites and settings. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from DNS query logs which could theoretically contain malicious content, though it is used here for diagnostic purposes.
- Ingestion points: DNS query logs retrieved through
homelab_adguardQueryLog(SKILL.md). - Boundary markers: No explicit delimiting or instruction-filtering markers are defined for the log data content.
- Capability inventory: The skill can restart services, modify configurations, and search logs using tools such as
homelab_serviceRestartandhomelab_adguardFilters. - Sanitization: No specific domain sanitization process is referenced for handling log entries.
- [EXTERNAL_DOWNLOADS]: The skill references DNS filter lists from established community repositories, including AdGuardTeam and StevenBlack on GitHub.
Audit Metadata