form-ux-design
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No patterns for bypassing safety guidelines or overriding agent behavior were detected.- [DATA_EXFILTRATION]: No sensitive file access or unauthorized network exfiltration patterns were found. Form data is handled through standard relative API endpoints.- [REMOTE_CODE_EXECUTION]: No dynamic code execution or remote script loading was identified.- [INDIRECT_PROMPT_INJECTION]: The skill identifies form inputs as a potential attack surface and provides robust remediation guidance.
- Ingestion points: Untrusted user input enters via web and mobile forms (documented in
references/form-validation.md). - Boundary markers: Instructions explicitly recommend validation and sanitization before processing.
- Capability inventory: Data is submitted to backend services via fetch/XHR.
- Sanitization: Recommends XSS prevention (
htmlspecialchars,escapeHtml) and parameterized SQL queries to mitigate risks.- [EXTERNAL_DOWNLOADS]: Mentions well-known libraries (Tom Select, Flatpickr, SweetAlert2) for UI enhancement without using suspicious or unverified sources.- [COMMAND_EXECUTION]: No dangerous shell commands or privilege escalation attempts were found.
Audit Metadata