The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to audit and transform untrusted MSBuild project files, which introduces a potential vector for malicious data to influence the agent's behavior.
Ingestion points: The skill ingests various XML-based build files, including .csproj, .vbproj, .fsproj, .props, and .targets files (referenced in SKILL.md).
Boundary markers: There are no explicit instructions or delimiters provided to the agent to distinguish between the file content and its own instructions, nor is there a directive to ignore embedded natural language instructions within the build files (e.g., in XML comments).
Capability inventory: The agent typically uses file-reading tools to access the code and may use file-writing or shell-execution tools to apply the recommended fixes listed in the catalog.
Sanitization: The skill does not define any sanitization or validation logic for the content of the files it analyzes.

msbuild-antipatterns