ai-test-generator
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests acceptance criteria from external YAML files and interpolates them directly into LLM prompts without sanitization or boundary markers. This creates a surface for indirect prompt injection where a malicious specification could influence the generated test code.
- Ingestion points: acceptance_criteria field in input YAML files (documented in SKILL.md Step 2).
- Boundary markers: None used in the prompt template.
- Capability inventory: The skill has file-write capabilities to the 'tests/generated/' directory via the save_test function in references/generation-and-scoring.md.
- Sanitization: No explicit validation or escaping of input criteria before prompt interpolation.
- [COMMAND_EXECUTION]: The skill executes shell commands 'npx tsc', 'python -m py_compile', and 'node --check' to validate the generated code. These commands are limited to syntax and compilation checks and do not involve execution of the generated test logic.
Audit Metadata