The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for the agent to review, build, and test C# and VB source code. This creates an attack surface for indirect prompt injection, where malicious instructions could be embedded in code comments, string literals, or metadata within the files provided by the user.
Ingestion points: User-provided source code and repository configuration files ingested during the implementation workflow (SKILL.md).
Boundary markers: The skill does not specify the use of delimiters or explicit instructions for the agent to ignore prompts embedded in external source data.
Capability inventory: The workflow involves file system operations (writing project files) and the execution of test harnesses to verify analyzer behavior.
Sanitization: No specific sanitization or filtering of external content is mentioned beyond standard Roslyn semantic analysis.
[EXTERNAL_DOWNLOADS]: The skill references configuration schemas and documentation templates from established development platforms.
Fetches a JSON schema for GitVersioning from the official .NET GitHub repository to configure project versioning.
Includes placeholder help links for diagnostic documentation intended for deployment to GitHub Pages.
[SAFE]: The implementation boilerplate for analyzers, source generators, and testing utilities follows official Roslyn patterns and utilizes reputable Microsoft libraries for code analysis.

roslyn-analyzers