agent-test-generator
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it analyzes and executes code derived from external sources.\n
- Ingestion points: Source code is read from local paths specified in the workspace to generate corresponding test suites.\n
- Boundary markers: The instructions do not include delimiters or specific instructions to disregard malicious commands that might be embedded in the code being analyzed.\n
- Capability inventory: The agent has access to the
Bashtool for executing shell commands andWrite/Edittools for file system modification.\n - Sanitization: The skill mandates the execution of generated tests without mentioning validation or sandboxing steps to mitigate potential malicious payloads in the source or generated tests.\n- [COMMAND_EXECUTION]: The skill explicitly uses the shell to perform its core tasks.\n
- Evidence: The
Bashtool is listed as a primary integration point for executing tests and measuring coverage.\n - Purpose: Shell access is required to run the Red-Green-Refactor cycle and ensure that the generated tests are functional and comprehensive.
Audit Metadata