unit-testing-test-generate
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
subprocess.runmethod to execute test commands and coverage tools. This execution path could be abused if thetest_commandvariable is influenced by untrusted input.\n- [DATA_EXFILTRATION]: The skill accesses the local filesystem to read source code for analysis using theopen()function. While necessary for its purpose, it allows reading of any file the agent has permissions for if the file path is manipulated.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted source code and possesses execution capabilities.\n - Ingestion points: The
_analyze_pythonmethod inSKILL.mdreads source code files directly from the filesystem.\n - Boundary markers: No explicit delimiters or warnings to ignore instructions within the source code are implemented in the prompt generation logic.\n
- Capability inventory: The skill can read local files (
open) and execute system commands (subprocess.run).\n - Sanitization: There is no evidence of path validation or command sanitization to prevent directory traversal or command injection.
Audit Metadata