azure-app-onboard-prereq
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- Access to Subscription Information: The skill utilizes standard Azure CLI commands (
az account show) to identify the active subscription and tenant. This information is used to populate session context files stored locally, ensuring the agent targets the correct Azure environment.\n- User-Gated Command Execution: While the skill has the capability to run project-specific build or installation commands (such asnpm installorpip install), these are strictly limited to remediation scenarios and are always preceded by an explicit user consent prompt for each command.\n- Static Analysis of Repository Content: To assess deployment feasibility, the skill scans local project files and manifests. This includes checking for hardcoded secrets, misconfigured environment variables, and deprecated frameworks, which are then reported to the user as actionable security or configuration findings.\n- Managed Code Modification: The skill can assist in remediating deployment blockers by suggesting or applying code changes. This functionality is restricted to specific deployment-related fixes and includes re-evaluation steps to ensure the changes align with Azure best practices.\n- Evaluation of Untrusted Repository Data: Since the skill processes information directly from the user's repository, it inherently handles untrusted data. It mitigates risks by focusing on static analysis and providing the user with visibility into all findings before proceeding with further steps.
Audit Metadata