network-health-check
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from a UniFi controller, which presents a surface for indirect prompt injection.
- Ingestion points: The skill ingests untrusted data from the UniFi controller via
unifi_list_alarms(specifically themsgfield) andunifi_list_devices(device names and status strings) as described inSKILL.md. - Boundary markers: No specific delimiters or boundary markers (such as XML tags or explicit instructions to the agent to ignore embedded commands) are used to wrap the ingested network data within the final report generation.
- Capability inventory: The tools available to the skill, as listed in
SKILL.mdand invoked viaunifi_batch, are restricted to diagnostic and information gathering functions (e.g.,unifi_get_system_info,unifi_get_network_health,unifi_list_devices,unifi_list_alarms). - Sanitization: The instructions do not specify any sanitization, escaping, or validation of the external alarm or device content before it is summarized for the user report.
Audit Metadata