dotnet-build-analysis

Fail

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides commands to install the Azure Artifacts Credential Provider by downloading and executing shell scripts from a Microsoft-owned domain. While the source is trusted, the execution pattern involves piping remote content directly to a shell or subshell. * Evidence: 'sh -c "$(curl -fsSL https://aka.ms/install-artifacts-credprovider.sh)"' in SKILL.md. * Evidence: 'iex "& { $(irm https://aka.ms/install-artifacts-credprovider.ps1) }"' in SKILL.md.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to interpret MSBuild and NuGet build logs, which are potentially untrusted data sources that could contain malicious instructions. * Ingestion points: MSBuild, NuGet, and Roslyn Analyzer logs processed throughout the skill documentation. * Boundary markers: No specific delimiters or instructions to ignore embedded content are provided. * Capability inventory: The skill suggests executing shell commands (dotnet, curl, ls) based on findings in the build logs. * Sanitization: No sanitization or validation of log input is mentioned.
  • [CREDENTIALS_UNSAFE]: The skill recommends using a NuGet CLI flag that stores sensitive credentials in plain text. * Evidence: 'dotnet nuget update source MyCompany --username az --password $PAT --store-password-in-clear-text' in SKILL.md.
Recommendations
  • HIGH: Downloads and executes remote code from: https://aka.ms/install-artifacts-credprovider.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 17, 2026, 07:31 AM
Security Audit — agent-trust-hub — dotnet-build-analysis