The Agent Skills Directory

[Authentication Security]: The skill emphasizes the use of DefaultAzureCredential for authentication. This is a recommended security practice as it avoids hardcoding secrets and leverages Azure's managed identity systems, reducing the risk of credential exposure.
[Secure Package Management]: The instructions use standard package management commands to fetch official Microsoft libraries from the NuGet registry. This ensures that the skill uses authentic and supported dependencies within a verified supply chain.
[Indirect Prompt Injection Surface]: The skill includes patterns for processing external message data, which creates a potential surface for indirect prompt injection. This is a standard consideration for applications handling external inputs.
Ingestion points: Message bodies are read in the Receive Messages and Background Processing sections of SKILL.md (e.g., message.Body.ToString()).
Boundary markers: No specific delimiters or instructions to ignore embedded commands are included when handling received message content in these basic examples.
Capability inventory: The skill demonstrates capabilities for sending messages and performing administrative actions like queue/topic creation and deletion via ServiceBusAdministrationClient in SKILL.md.
Sanitization: No explicit sanitization or validation of the received message body is shown in the simplified code examples, which is common for library documentation but warrants developer awareness when building production systems.

azure-servicebus-dotnet