test-gap-analysis
Warn
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies and executes test commands directly from project manifests (e.g., package.json, .csproj). This behavior allows the execution of arbitrary shell commands if the analyzed project manifest is malicious.\n
- Evidence in SKILL.md: "Run the narrowest existing test command once. Choose it from the project manifest; Microsoft.Testing.Platform executables may require
dotnet run."\n- [DYNAMIC_EXECUTION]: The skill performs runtime modification of production source code ("mutations") and executes the test suite against the modified code. This represents a significant execution surface where the agent's behavior is driven by dynamically altered code.\n - Evidence in SKILL.md: "Apply one candidate and confirm the diff changes exactly one intended expression. Run the narrowest covering test..."\n- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect injection via malicious project data. It ingests untrusted production files, test files, and manifests to determine its execution logic and commands.\n
- Ingestion points: Production and test files, project manifests (SKILL.md).\n
- Boundary markers: Not explicitly defined for external data processing.\n
- Capability inventory: File reading, file writing (mutations), and command execution (SKILL.md).\n
- Sanitization: The skill does not describe validation or sanitization of commands extracted from manifests before execution.\n- [EXTERNAL_DOWNLOADS]: The skill references an external test-analysis-extensions catalog that can be provided by the caller or runtime environment. This allows the agent to load and follow language-specific rules from potentially untrusted paths.\n
- Evidence in SKILL.md: "...read the matching language file relative to the caller-provided or runtime-listed test-analysis-extensions catalog..."
Audit Metadata