tdd-workflow
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform local command execution, specifically 'npm test' and 'npm run test:coverage', as part of the development lifecycle.
- [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface (Category 8) by utilizing untrusted user-defined 'User Journeys' to guide code generation and subsequent execution.
- Ingestion points: User journeys provided in 'Step 1' of the workflow (SKILL.md).
- Boundary markers: No delimiters are specified to isolate user input from the agent's instructions.
- Capability inventory: The skill uses 'npm test' (SKILL.md) to execute code derived from user input.
- Sanitization: No input validation or sanitization of user-provided journeys is mentioned.
Audit Metadata