The Agent Skills Directory

[Handling of Sensitive Data]: The skill provides code snippets that demonstrate how to retrieve and print subscription keys and uses a placeholder password in examples. While these are common in documentation, they represent a potential security consideration regarding the exposure of secrets in console logs or development environments.
[Indirect Prompt Injection Surface]: The skill includes functionality to import API definitions from external OpenAPI specifications provided via local files or remote URLs. This introduces a surface where malformed or malicious data in external specifications could attempt to influence the agent's context.
Ingestion points: Importing API content in references/apis-operations.md via file reading or URL links.
Boundary markers: The instructions do not define specific delimiters to isolate external data from the prompt context.
Capability inventory: The skill possesses extensive capabilities to create, modify, and delete Azure resources, including the application of API policies.
Sanitization: Validation is primarily handled by the Azure Resource Manager service layer at runtime.
[External Resource References]: The skill references external domains such as petstore.swagger.io for sample API imports and official Azure documentation. These references target well-known and established services for development and learning purposes.

azure-mgmt-apimanagement-dotnet