The Agent Skills Directory

Repository Overlay Ingestion: The skill is designed to read and follow instructions from a file located at .agents/skill-overlays/dotnet-test/writing-mstest-tests.md within the repository. This mechanism allows external repository content to influence agent behavior and naming conventions, which could potentially lead to indirect prompt injection if the repository contains untrusted or malicious instructions.
Ingestion points: The skill reads the repository overlay file and processes user-provided production and test code.
Boundary markers: The skill checks for specific frontmatter fields (core, binding-revision, and mode) in the overlay file to verify compatibility.
Capability inventory: The skill performs file system reads and can trigger dotnet test command execution.
Sanitization: The skill performs basic validation of the overlay's metadata before applying its instructions.
Command Execution via Test Runner: The workflow includes the execution of the dotnet test command. Running tests naturally involves compiling and executing code within the environment, which is a standard security consideration when working with unknown or third-party test suites.
Standard Package Management: The skill provides examples for adding MSTest and MSTest.Sdk packages. These are standard dependencies for .NET development projects and are retrieved from well-known package registries.

writing-mstest-tests