test-generator
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests source code from files to generate tests, which can expose the agent to malicious instructions embedded within the codebase.\n
- Ingestion points: Step 1 of the process in SKILL.md involves reading source files.\n
- Boundary markers: The instructions lack delimiters or specific guidance to ignore embedded instructions in data.\n
- Capability inventory: Filesystem read/write and shell command execution via a test runner.\n
- Sanitization: No content sanitization is described before processing the source files.\n- [COMMAND_EXECUTION]: The verification process instructs the agent to execute a shell command to run the project's test suite.\n
- Evidence: The 'Verification Loop' in SKILL.md directs the agent to 'run them with the project's test runner'.\n- [DYNAMIC_EXECUTION]: The skill generates executable code and immediately triggers its execution through a test runner.\n
- Evidence: The workflow in SKILL.md generates test files and follows up with an automated verification loop that executes the generated code.
Audit Metadata