openui
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for using the
@openuidev/clitool to perform administrative tasks such as scaffolding new projects (create), generating API keys, deploying applications, and submitting feedback. - [EXTERNAL_DOWNLOADS]: The instructions rely on several libraries and tools from the
@openuidevscope on NPM, which are vendor resources associated with the skill's author. It also references well-known services like Vercel and OpenAI. - [INDIRECT_PROMPT_INJECTION]: The skill addresses potential injection risks when processing untrusted data from external sources or model-generated code. It mitigates these by instructing the agent to treat repository content as non-authoritative reference data and recommending the use of sandboxed iframes for rendering HTML artifacts.
- [DYNAMIC_EXECUTION]: At its core, the skill involves parsing and rendering a custom, model-generated language (OpenUI Lang). The provided documentation includes comprehensive guides on validation, error handling, and reliability monitoring to manage the risks of executing dynamically generated UI structures.
Audit Metadata