gaokao-liberal-arts-tutor
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION] (SAFE): The instructions in SKILL.md are strictly pedagogical, focusing on heuristic teaching methods and academic guidance for History, Geography, and Politics. No bypass patterns or safety override attempts were detected.
- [EXTERNAL_DOWNLOADS] (LOW): The README.md suggests a manual installation process via 'git clone' from an unverified GitHub repository ('flysheep-ai/education-skills'). While this source is not on the pre-approved trusted list, the action is documented for the user and not performed automatically by the skill's code.
- [DATA_EXFILTRATION] (SAFE): There are no references to sensitive file paths (~/.ssh, .env, etc.) and no network calls (curl, fetch) present in the skill files.
- [COMMAND_EXECUTION] (SAFE): The skill does not include any scripts (.sh, .py, .js) or instructions that would cause the agent to execute shell commands or spawn subprocesses.
Audit Metadata