test-generation

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the command line to run generated test suites.
  • Evidence: Step 5 explicitly invokes npm test to execute the generated test files.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted feature requirements to generate and execute code.
  • Ingestion points: Step 1 accepts user-defined 'Feature' and 'Behaviors' text inputs.
  • Boundary markers: Absent. There are no explicit instructions to the agent to disregard instructions embedded within the user-provided feature requirements.
  • Capability inventory: The skill can write files and execute them using npm test (Category 10 behavior).
  • Sanitization: Absent. No input validation or code sanitization logic is defined for the requirements before they are incorporated into the generated test code.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 12:50 PM