The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes dotnet build --no-restore and dotnet list package --vulnerable to evaluate project health. While these are standard for the stated purpose, executing build commands on untrusted codebases can trigger malicious code via project-file hooks such as MSBuild targets.
[DYNAMIC_EXECUTION]: The skill triggers runtime compilation of .NET source code using the dotnet build command to assess project build health.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted content from the target codebase, creating a surface where malicious instructions in the source could influence the assessment results.
Ingestion points: The skill reads project source code, project graphs, and documentation files (README, XML docs, and CLAUDE.md) using various MCP tools and file system operations.
Boundary markers: No explicit delimiters or isolation instructions are provided to separate analyzed project content from the agent's instructions.
Capability inventory: The skill allows the agent to execute shell commands and read project-wide files, which could be abused if an injection is successful.
Sanitization: The skill implements a triage gate that prioritizes summaries and high-confidence tool findings, providing some mitigation, but it does not sanitize instructions embedded within the codebase data.

health-check