money-quality
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute standard development CLI tools including
npx,npm,python, andgofor linting, type checking, and running test suites. These are legitimate operations for a quality engineering skill.\n- [PROMPT_INJECTION]: The skill defines a 'Standard startup' sequence to initialize the agent's workflow, loading project-specific context and frameworks. This serves as a structured instructional setup for the agent.\n- [PROMPT_INJECTION]: The skill reviews untrusted code diffs and staged changes, creating an indirect prompt injection surface. While it lacks explicit boundary markers to separate instructions from code content, this is inherent to the task of code review.\n - Ingestion points: Local project files and code diffs (SKILL.md).\n
- Boundary markers: None specified.\n
- Capability inventory: Shell command execution (testing, linting) and file reading.\n
- Sanitization: No sanitization of the code being reviewed is described.
Audit Metadata