ai-phd-expert
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to provide academic and technical assistance for AI/ML tasks and does not contain any malicious code or instructions.- [PROMPT_INJECTION]: No evidence was found of attempts to bypass safety filters, override system instructions, or extract system prompts.- [DATA_EXPOSURE]: The skill does not access sensitive file paths (like SSH keys or env files) and contains no hardcoded credentials.- [REMOTE_CODE_EXECUTION]: While the skill assists with code generation, it does not instruct the agent to download or execute remote scripts from untrusted sources.- [COMMAND_EXECUTION]: The instruction to save summaries to .md files is a benign productivity feature and does not involve the execution of arbitrary or elevated system commands.
Audit Metadata