ops-inspector
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for operational inspection and health checks, using pre-defined platform tools to retrieve resource metrics and logs. All referenced documentation and console links point to legitimate vendor infrastructure.
- [INDIRECT_PROMPT_INJECTION]: The skill processes environment logs which could potentially contain untrusted content (Ingestion points:
queryLogsandqueryFunctionslogs inSKILL.md). While specific boundary markers are not explicitly defined, the agent's actions are restricted to diagnostic reporting and information retrieval (Capability inventory: CloudBase resource metadata and metrics), posing no significant escalation risk (Sanitization: None described). - [CREDENTIALS_UNSAFE]: The instructions include specific security guidance to prevent the storage or transmission of secrets, advising users to manage authentication through established environment-binding tools instead of manual entry.
- [EXTERNAL_DOWNLOADS]: The skill explicitly forbids the HTTP-fetching of remote content into the agent context, mandating the use of local sibling skills and local reference files.
Audit Metadata