microsoft-foundry
Pass
Audited by Gen Agent Trust Hub on Apr 9, 2026
Risk Level: SAFE
Full Analysis
- External Resource Integration: The skill fetches starter samples and templates from official Microsoft GitHub repositories (such as
microsoft-foundry/foundry-samples) to assist in agent creation. These are trusted sources used to provide developer quick-starts and are verified for integrity during the onboarding process. - Automated Workflow Execution: Utilizes standard command-line tools, including the Azure CLI, Docker, and the Azure Developer CLI (
azd), to automate resource provisioning and container management tasks. These operations are conducted within the user's authenticated Azure context. - Evaluation Data Harvesting Surface: Includes a structured pipeline for harvesting production traces from Application Insights to improve agent performance. While this involves processing data from external logs, the skill implements a mandatory human-in-the-loop curation step to ensure data quality and security before inclusion in evaluation sets.
- Dynamic Scripting and Command Processing: Uses local scripts to perform complex operations like ranking model capacity across Azure regions. These scripts process data using standard utilities such as
jqand the Python interpreter, facilitating efficient management of Azure resources. - Authentication and Secret Management: Strictly adheres to Azure security best practices by recommending the use of managed identities and environment-aware credential resolution rather than hardcoding sensitive information.
Audit Metadata