The Agent Skills Directory

Command Execution: The skill utilizes the dotnet new CLI to perform search, list, and inspection tasks. This is consistent with the skill's purpose and uses standard development tools provided by the vendor.
External Downloads: The instructions include searching NuGet.org and installing template packages via dotnet new install. While these operations fetch remote content, they are standard procedures within the .NET ecosystem for managing project templates.
Indirect Prompt Injection: The skill processes data from external sources that could contain untrusted instructions. (1) Ingestion points: User-provided intent and CLI output from dotnet new commands. (2) Boundary markers: Not explicitly defined in the skill body. (3) Capability inventory: Execution of dotnet new commands in SKILL.md. (4) Sanitization: No specific filtering or escaping is defined. This surface is a potential consideration when the agent interacts with third-party template metadata.

template-discovery