cost-optimization
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface by ingesting untrusted data from external Azure environments (metrics, cost reports, and resource metadata) and utilizing it to execute potentially destructive capabilities.
- Ingestion points: The skill ingests 30-day utilization metrics (CPU, memory, DTU) and Azure Cost Management data to identify optimization targets (documented in SKILL.md sections 'Right-Sizing' and 'Cost Analysis Process').
- Boundary markers: There are no explicit boundary markers or instructions to ignore embedded prompts within the processed Azure data.
- Capability inventory: The skill possesses capabilities to list and delete Azure resources (e.g.,
az disk list,az disk delete), discover tools viatool_search_tool_regex, and invoke pricing tools (documented in SKILL.md sections 'Eliminate Waste' and 'Cost Analysis Process'). - Sanitization: No sanitization or validation logic is defined for the external resource IDs or metric values before they are used in shell commands or tool arguments.
Audit Metadata