grade-tests
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFE
Full Analysis
- Data Ingestion for Analysis: The skill processes test method bodies, file paths, and production code retrieved from the workspace. This data ingestion is necessary for the skill's primary function of grading code quality. While processing untrusted data (such as code from a pull request) is a common pattern in analysis tools, the skill focuses on evaluation and reporting without executing the content.
- Modular Skill Dependency: The instructions include calls to an external skill,
test-analysis-extensions, to load language-specific framework rules (e.g., .NET, Python, Go). This design pattern ensures that the agent uses specialized logic for different environments, which is a standard approach for polyglot tools. - Strict Input Validation: The skill implements a 'Step 0' validation check that requires explicit scopes (file paths or method names) before proceeding. This safeguard helps prevent the agent from performing unintended wide-scale operations on the workspace when the user's request is ambiguous.
- Analytical Focus: The skill's workflow is centered on generating markdown reports based on static code patterns. It specifically instructs the agent not to attempt fixing failures or measuring coverage, which maintains a clear boundary around its intended capabilities.
Audit Metadata