balise-trustworthy-flows
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSNO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements strong safety principles, including instructions to escalate serious risks (privacy, safety, legal) to specialists and to never declare a system legal or compliant based on static artifacts.
- [EXTERNAL_DOWNLOADS]: References official documentation from trusted regulatory bodies such as the FTC, OECD, and EDPB for guidance on identifying deceptive design mechanisms.
- [PROMPT_INJECTION]: The skill's analysis of untrusted artifacts constitutes an indirect prompt injection surface. The skill mitigates this through: 1) Ingestion points: Figma files, screenshots, and frontend code; 2) Boundary markers: Mandatory evidence labels (Observed, Runtime Verified); 3) Capability inventory: File reading and labeled editing; 4) Sanitization: Requirements to trace visible behavior to technical evidence.
- [NO_CODE]: The skill consists entirely of instructional Markdown documentation and does not contain any executable scripts or binary files.
Audit Metadata