ml-azureml-adf-automation
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious behavior detected. The skill provides legitimate automation guidelines for Microsoft Azure services.
- [EXTERNAL_DOWNLOADS]: Recommends installing standard Azure SDK packages (azure-ai-ml, azure-identity) and a library constraint (marshmallow) via pip.
- [COMMAND_EXECUTION]: Provides scripts using Azure CLI and PowerShell for resource management, which is consistent with the skill's purpose.
- [DATA_EXFILTRATION]: Makes a request to https://api.ipify.org to obtain the agent's public IP address for configuring temporary storage firewall rules. This is a standard practice for secure CI access to private storage.
- [CREDENTIALS_UNSAFE]: Correctly utilizes AzureCliCredential and Managed Service Identity (MSI) for authentication, avoiding the use of hardcoded credentials.
Audit Metadata