skills/dotnet/skills/eval-performance/Gen Agent Trust Hub

eval-performance

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Build Diagnostic Command Execution: The skill provides instructions for executing standard MSBuild diagnostic commands, such as dotnet msbuild -pp for preprocessing and dotnet msbuild with performance summaries. These are legitimate administrative and developer tools used to profile build performance and do not exhibit suspicious behavior.
  • Log Analysis and Text Processing: The skill utilizes grep to extract timing information and evaluation events from build logs (full.log). This is a standard practice for performance analysis in software development environments.
  • Indirect Prompt Injection Surface: As a diagnostic skill, it is designed to ingest and process build artifacts like binlogs and log files. While these files are external data sources, the skill treats them as diagnostic input for performance analysis, which is its primary intended function. The ingestion of these developer-controlled artifacts is a standard part of the MSBuild profiling workflow.
  • Diagnostic Tool Integration: The skill references the Microsoft.AITools.BinlogMcp tool (via the binlog MCP namespace) for structured analysis of build evaluations. This integration allows for precise querying of build properties and durations, following best practices for automated build diagnostics.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:14 AM
Security Audit — agent-trust-hub — eval-performance