The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes C# code and service definitions within a workspace. This creates an attack surface where malicious instructions could be embedded in the source code. * Ingestion points: Analyzes source code, service boundaries, and APIs in the workspace. * Boundary markers: No explicit delimiters or warnings are defined to isolate untrusted code content from the agent instructions. * Capability inventory: The skill operates with workspace-write permissions, allowing it to modify files in the environment based on analyzed content. * Sanitization: No specific validation, escaping, or filtering of the processed code content is mentioned.

csharp-developer