accessibility-fix

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow fetches and parses external web pages the user provides (Step 1: “Fetch and Parse Page”), and that fetched outsider-authored HTML/text is ingested into the agent’s LLM context for auditing.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches user-provided pages at runtime (e.g., https://example.com/about) and related resources (/nav.plain.html, /footer.plain.html, and site styles.css) and then uses that fetched HTML/CSS as the primary input to generate prompts/fixes, so those runtime fetches can directly control agent output.