onesignal-iam-html-composer
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is developed by 'OneSignal' and focuses on its primary business function of creating and refining in-app message content. It correctly utilizes official vendor tools (
apply_to_iam_composer,generate_brand_kit_from_website) and JS APIs (OneSignalIamApi).- [SAFE]: A human-in-the-loop security control is enforced through the 'Summary and Approval' workflow, which requires the agent to present proposed changes and obtain explicit user consent before any HTML code is delivered or applied to the platform.- [SAFE]: The instructions include a mandatory 'Self-Check' procedure where the agent must verify the rendering soundness and API correctness of the generated HTML, serving as a safeguard against improper code generation.- [SAFE]: While the skill processes user-supplied data (prompts, URLs, and pasted HTML), the potential for indirect prompt injection is mitigated by the restricted operational scope, the sandboxed execution environment (WebView), and the requirement for user approval before changes take effect.
Audit Metadata