azuresql-db-testing

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [External Downloads]: The skill instructs the setup of integration tests that fetch a container image from a private registry (sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io). This is a standard requirement for accessing private preview features and is managed via standard Docker authentication mechanisms.
  • [Command Execution]: The provided test recipes utilize wait-for-ready strategies that execute commands, such as sqlcmd, inside the ephemeral container to ensure the database engine has initialized before the tests begin. This execution is scoped to the containerized environment created for the duration of the test run.
  • [Indirect Prompt Injection Surface]: The skill serves as a template generator for integration tests, creating code that interacts with container APIs and database engines based on user input.
  • Ingestion points: The skill processes user requests for specific testing frameworks and database requirements to generate appropriate fixture code.
  • Boundary markers: The skill uses clear validation rules and per-language snippets to guide the agent in generating structured test code.
  • Capability inventory: The generated code includes starting/stopping containers, mapped port retrieval, and SQL database provisioning (CREATE DATABASE).
  • Sanitization: The snippets utilize hardcoded placeholders and constants for sensitive values like passwords and database names within the test environment, minimizing the risk of user-supplied content influencing the execution logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 10:32 AM
Security Audit — agent-trust-hub — azuresql-db-testing