wix-docs

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

• Third-party content exposure detected (high risk: 0.95). The required runtime workflow uses curl to fetch Wix documentation pages from dev.wix.com (public web content) and appends .md, so the LLM ingests outsider-authored free text from those fetched markdown pages via the “find the page, then read it” path.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs sending JavaScript functions at runtime to https://mcp.wix.com/api/code-mode/search (POSTing a "code" async function) which the service executes in a sandbox and returns JSON, meaning the agent depends on executing remote-submitted code to produce/shape results.