audit-accessibility

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text is fetched at runtime from public sources via the MCP tool calls to user-firecrawl (web pages scraped as formats: ["markdown"]), and that scraped markdown can be injected into the agent/LLM context for “research” summarization.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill injects and executes remote JavaScript at runtime from the CDN URL https://cdnjs.cloudflare.com/ajax/libs/axe-core/4.10.2/axe.min.js (script.src is appended and axe.run() is called), so this external URL is fetched and executed during runtime.