learning-diagnostic-assessment
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [PROMPT_INJECTION] (SAFE): The documentation contains standard instructional language for educational assessment design and does not include any attempts to bypass safety filters or override system instructions.
- [DATA_EXFILTRATION] (SAFE): The script does not utilize any network libraries (like requests or urllib) or perform any operations that could transmit data externally. It has no access to sensitive file paths.
- [REMOTE_CODE_EXECUTION] (SAFE): There is no evidence of dynamic code execution (eval, exec) or fetching of remote scripts. The skill only performs basic arithmetic for item distribution.
- [COMMAND_EXECUTION] (SAFE): The skill does not use subprocess or os.system to interact with the underlying host shell. The CLI examples in the markdown are illustrative and not backed by dangerous code in the Python implementation.
- [INDIRECT_PROMPT_INJECTION] (SAFE): While the skill accepts user input for 'domain' and 'skill_areas', it lacks any sinks (such as shell commands or file writes) that would allow an attacker to exploit this ingestion point.
Audit Metadata