az-cost-optimize
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Azure CLI (az) and specialized Azure MCP tools (such as azmcp-subscription-list and azmcp-resource-group-list) to query the cloud environment for resource discovery and metrics collection. This behavior is consistent with the stated administrative purpose.
- [EXTERNAL_DOWNLOADS]: The workflow references the official Azure pricing portal (azure.microsoft.com/pricing/) to fetch cost data. This is a reference to a well-known service and is documented neutrally.
- [DATA_EXFILTRATION]: Configuration details and usage patterns are extracted from Azure and transmitted to GitHub via the creation of issues. This cross-platform data transfer is the primary function of the skill for reporting optimization recommendations.
- [PROMPT_INJECTION]: The skill processes untrusted external data which creates a surface for indirect prompt injection.
- Ingestion points: The skill reads local Infrastructure-as-Code (IaC) files (.bicep, .tf, .json) via file_search and ingests monitoring log data via azmcp-monitor-log-query.
- Boundary markers: No specific delimiters or instructions to ignore embedded commands are used when processing these inputs.
- Capability inventory: The agent has the capability to create GitHub issues and execute Azure management commands.
- Sanitization: No explicit sanitization or validation of the ingested external content is mentioned.
- Mitigation: The risk of automated malicious actions is mitigated by a mandatory user confirmation step (Step 5) that requires human approval before issues are created or implementation starts.
Audit Metadata