test-generator
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted source code to generate executable test files.
- Ingestion points: Reads project source files to understand logic and dependencies (
SKILL.md). - Boundary markers: No delimiters or instructions are provided to distinguish between source code and potential malicious instructions embedded within it.
- Capability inventory: The skill creates new files adjacent to source files and executes the project's test runner (
SKILL.md). - Sanitization: There is no evidence of validation or sanitization of the source code content before it is used to generate test scripts.\n- [COMMAND_EXECUTION]: The skill requires the ability to execute the project's test runner in the local environment.
- Evidence: The
Verification LoopinSKILL.mdexplicitly instructs the agent to run the generated tests using the project's test runner up to three times to ensure they pass.
Audit Metadata