csharp-expert

Pass

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill operates as a dispatcher, processing user input and repository metadata to route requests to specific specialist tools.
  • Ingestion points: The skill ingests user prompts and repository files, including MSBuild project files (.csproj), diagnostic logs (.binlog), and source manifests, to determine the appropriate specialist.
  • Boundary markers: Instructions emphasize scoping exploration to "load-bearing" files, though they do not explicitly mandate the use of delimiters when passing this context to sub-skills.
  • Capability inventory: The skill is authorized to invoke other specialist skills via the skill tool and uses local repository navigation and diagnostic tools for analysis.
  • Sanitization: There are no explicit instructions to sanitize or escape data extracted from the repository before it is interpolated into tool calls.
  • [REMOTE_CODE_EXECUTION]: The skill includes instructions for the manual installation of individual skills from a remote source.
  • Evidence: The command skill-installer install https://github.com/dotnet/skills/... is provided as a fallback acquisition path.
  • Context: The source is an official repository from a trusted organization.
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for obtaining additional capabilities from a remote marketplace.
  • Evidence: The skill references github.com/dotnet/skills and provides host-level commands such as /plugin install <plugin>@dotnet-agent-skills.
  • Context: These references facilitate the acquisition of official .NET tools from a trusted source.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 11, 2026, 05:11 AM