syncfusion-vue-rich-text-editor
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation includes instructions to install various NPM packages from official registries, including
@syncfusion/ej2-vue-richtexteditorand themarkedlibrary. These are standard dependencies required for the editor's core functionality and markdown support. - [INDIRECT_PROMPT_INJECTION]: The skill implements an AI Assistant feature (
AIAssistantService) that interpolates user-controlled editor content into prompts sent to an external AI service. - Ingestion points: The editor's content area where untrusted user input is received (e.g.,
SKILL.md,references/editor-value.md). - Boundary markers: The provided code snippets in
references/ai-assistant.mddo not include delimiters or instruction-isolation techniques when concatenating user text into the AI prompt (args.prompt + args.text). - Capability inventory: The skill performs network operations to configured AI service endpoints via
fetchand can modify the editor's content based on the response usingaddAIPromptResponse. - Sanitization: The component includes robust sanitization features (
enableHtmlSanitizer,PasteCleanup) to prevent cross-site scripting (XSS), although these are not designed to filter semantic prompt injection attacks.
Audit Metadata