syncfusion-vue-inline-ai-assist

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The component processes user-selected text and document content to generate AI prompts. In some examples, this content is interpolated directly into prompts (e.g., using string replacement for placeholders like '{selection}'), which is a known surface for indirect prompt injection.
  • Ingestion points: SKILL.md and references/command-settings.md capture user data via window.getSelection() and innerText.
  • Boundary markers: Code examples do not demonstrate the use of delimiters or 'ignore' instructions to isolate untrusted data.
  • Capability inventory: The skill uses fetch for network operations and innerHTML for updating the document with AI responses.
  • Sanitization: No sanitization logic for user-provided text is included in the implementation examples.- [DYNAMIC_EXECUTION]: The skill demonstrates patterns where AI-generated responses are rendered directly into the DOM using innerHTML or via custom templates (editorTemplate, responseTemplate). This presents a risk of script execution (XSS) if the AI service returns malicious HTML content.- [EXTERNAL_DOWNLOADS]: The documentation instructs users to install official packages from the @syncfusion namespace on NPM. These are verified vendor resources.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 09:41 PM