building-html-canvases
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of documentation and instructional guidelines. No executable scripts, binaries, or active code are included in the skill files.
- [SAFE]: The instructions reinforce platform security controls, explicitly mentioning that remote scripts and dynamic imports are blocked and that external network requests require pre-defined origin allowlists.
- [SAFE]: Data access and external navigation are directed through the platform's internal bridge (
ph), which is the standard and audited method for agent communication within this environment. - [SAFE]: The recommended methods for theme handling and styling rely on standard browser APIs (
getComputedStyle) and internal CSS variables, posing no risk to the host environment.
Audit Metadata