study-strategy-selector
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected during the analysis of the instructions and metadata.
- [PROMPT_INJECTION]: The prompt instructions establish a professional academic persona and focus strictly on the educational task of recommending study strategies. There are no attempts to bypass safety filters, override system instructions, or extract system prompts.
- [DATA_EXFILTRATION]: The skill's input schema includes a field for 'student_profiles' which may contain sensitive academic data. However, the skill does not utilize any network tools (curl, wget) or file system operations that would allow this data to be exfiltrated. All processing is confined to generating a text response within the agent environment.
- [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or scripts. It operates entirely through natural language instructions and template interpolation.
- [REMOTE_CODE_EXECUTION]: There are no patterns indicating the download or execution of remote scripts, and no external dependencies are specified.
Audit Metadata