design-md

Warn

Audited by Snyk on Feb 27, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.80). The skill's required Retrieval & Networking steps instruct the agent to download and parse project HTML and screenshots from the Stitch MCP Server (e.g., using web_fetch or read_url_content on the screen's htmlCode.downloadUrl and screenshot.downloadUrl), which are user/project-provided assets that the agent must read and which can directly influence its synthesis and subsequent actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 27, 2026, 02:26 PM