The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several shell commands as part of its scaffolding and verification workflow.
Evidence: Instructions include dotnet restore, dotnet build, dotnet tool restore, dotnet run, and Entity Framework commands like dotnet ef migrations add and dotnet ef database update.
[EXTERNAL_DOWNLOADS]: The skill adds necessary NuGet packages and restores .NET tools from official registries.
Evidence: Workflow Step 1 and 5 instruct the agent to add package references (e.g., Microsoft.EntityFrameworkCore.Design) and restore local or global tools using dotnet tool restore.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes natural language requests and inspects existing project files to generate new code.
Ingestion points: Natural language scaffolding requests and local project files (.csproj, Program.cs, model classes).
Boundary markers: None explicitly defined to separate user input from system instructions.
Capability inventory: File writing, shell command execution via dotnet CLI, and local network request execution via .http files.
Sanitization: Relies on the agent's internal generative logic without explicit validation steps.

create-datadriven-aspnetcore