pi-web-ui
Pass
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, prompt injections, or obfuscation techniques were detected in the skill instructions or reference documentation.
- [COMMAND_EXECUTION]: The skill includes a
javascript_repltool and anArtifactsPanelthat enable the execution of JavaScript and HTML. This is a primary feature of the library and is implemented using aSandboxIframeto provide execution isolation within the browser. - [EXTERNAL_DOWNLOADS]: Includes a
createExtractDocumentToolcapable of fetching remote documents (PDF, DOCX, etc.) from external URLs for text extraction. It supports the use of a CORS proxy (e.g., corsproxy.io) to facilitate these requests from the browser. - [CREDENTIALS_UNSAFE]: The library provides components like
ProviderKeysStoreandApiKeyPromptDialogto manage AI provider API keys. These are stored locally in the user's browser via IndexedDB and are used to authenticate requests to LLM providers, which is standard functionality for a client-side AI interface. - [SAFE]: The skill exhibits an indirect prompt injection surface by processing external documents and URLs; however, it utilizes sandboxing for code execution and standard Lit component rendering, following security best practices for this category of tool.
Audit Metadata