monitoring
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
zillizCLI tool to perform monitoring operations. All commands (such aszilliz cluster describe,zilliz collection list, andzilliz database list) are scoped to retrieving metadata, status information, and metrics from the Zilliz Cloud environment. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data returned by the Zilliz CLI to generate summaries and tables for the user. This creates a data ingestion surface where external content (e.g., collection names or status descriptions) is processed by the agent.
- Ingestion points: Data from
zilliz collection list,zilliz collection get-stats, andzilliz collection get-load-state(SKILL.md). - Boundary markers: None explicitly defined in the skill instructions.
- Capability inventory: Limited to CLI-based status retrieval; no generic shell execution or arbitrary file-write capabilities are present in this skill.
- Sanitization: Not explicitly mentioned, however, the skill focuses on formatting specific database fields rather than processing free-text user content.
Audit Metadata