ui-planning-and-scaffolding
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions guiding the agent's interaction with the user. It contains no executable scripts, hardcoded credentials, or network-enabled commands.
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits a potential indirect injection surface as it analyzes existing project code to inform its scaffolding behavior. Evidence Chain: (1) Ingestion points: User responses and existing project files (e.g., .qml files). (2) Boundary markers: Not specified. (3) Capability inventory: File-system write operations for scaffolding. (4) Sanitization: Not specified. This behavior is inherent to the skill's primary purpose of contextual UI generation.
Audit Metadata