The Agent Skills Directory

[Integration with Trusted Services]: The skill references official Microsoft and OpenAI documentation and endpoints (such as azure.com and aka.ms). These resources are provided by the vendor and are appropriate for the migration task.
[Secure Authentication Practices]: Code examples and instructions prioritize the use of environment variables and Azure Entra ID (via DefaultAzureCredential) for authentication. This approach follows security best practices by avoiding hardcoded secrets and utilizing managed identities.
[Local Workspace Scanning]: The included scripts/detect_legacy.py utility allows for identifying migration targets through local regex-based analysis. This script operates strictly on the file system and does not perform network operations or execute untrusted code.
[Indirect Prompt Injection Surface]: As the skill is designed to analyze and modify user-provided source code, it possesses a surface for indirect prompt injection if the codebase contains adversarial strings. This is a common property of code-refactoring skills and is mitigated by the skill's specific purpose and model safety guardrails.
Ingestion points: The agent and the provided scanning script read contents from files within the local workspace.
Boundary markers: The skill relies on standard file structures and does not implement specific delimiters for code ingestion.
Capability inventory: Capabilities include reading and writing to the local working tree and executing the provided Python scanning script.
Sanitization: Content is analyzed for API patterns rather than sanitized for adversarial instructions, which is standard for migration tooling.

azure-openai-to-responses