gentle-teaching
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill contains no executable scripts (Python, JavaScript, shell) or configuration files that could be used to execute commands on a host system. It is purely an instructional framework.
- [PROMPT_INJECTION] (SAFE): No patterns of prompt injection, role-play bypass, or instruction-override techniques were found. In fact, the skill explicitly mandates 'Boundary Maintenance' and 'Explicit Parameters' to ensure the AI remains in a controlled teaching mode.
- [EXTERNAL_DOWNLOADS] (SAFE): There are no references to external URLs, package managers (npm, pip), or remote script execution (curl|bash).
- [DATA_EXFILTRATION] (SAFE): The skill lacks any network-capable code or logic that could read sensitive files and transmit them to an external domain.
- [CREDENTIALS_UNSAFE] (SAFE): No hardcoded API keys, tokens, or secrets were detected in the metadata or the markdown body.
Audit Metadata