dotnet-build-analysis

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides commands to download and execute the Azure Artifacts Credential Provider directly from Microsoft's official 'aka.ms' domain.
  • Evidence: Commands include 'iex "& { $(irm https://aka.ms/install-artifacts-credprovider.ps1) }"' for Windows and 'sh -c "$(curl -fsSL https://aka.ms/install-artifacts-credprovider.sh)"' for macOS/Linux.
  • [COMMAND_EXECUTION]: Provides templates for common development and build system commands.
  • Evidence: Includes usage of 'dotnet nuget', 'dotnet workload', 'ls', and 'docker run' to diagnose environment and configuration issues.
  • [PROMPT_INJECTION]: The skill guides an agent to parse and act upon build output (logs). Because build logs can include content derived from source code or external project files, they represent a surface for indirect prompt injection.
  • Ingestion points: MSBuild output strings, error messages (CS, MSB, NU, IDE, CA), and file paths from 'SKILL.md'.
  • Boundary markers: None specified for separating log data from agent instructions.
  • Capability inventory: The agent is encouraged to execute 'dotnet' CLI commands and modify source code/configuration based on log interpretation.
  • Sanitization: No explicit sanitization or validation of the build log content is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 01:12 PM