find-top-three
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external data provided by the user, such as resumes and decision records, and incorporates findings into files written to the agent's workspace. This pattern creates a surface for indirect prompt injection.\n
- Ingestion points: Processes user-provided resumes, financial records, and data screenshots (modeling-framework.md).\n
- Boundary markers: Instructions explicitly require separating 'Confirmed Facts' from 'User Interpretations' and 'Model Hypotheses' (modeling-framework.md).\n
- Capability inventory: The skill instructs the agent to create and update markdown files in the local filesystem (output-templates.md).\n
- Sanitization: No explicit sanitization or escaping of the ingested content is defined before it is written to the output files.\n- [SAFE]: No hardcoded credentials, malicious network operations, or obfuscated payloads were detected.\n- [SAFE]: The skill follows privacy best practices by requiring user consent before persisting information and advising against storing sensitive personal details.
Audit Metadata