The single-call assessment uses Gemini's url_context tool to fetch the live web page and evaluate attributed response segments in one prompt, extracting survived, lost, and distorted claims to compute the URL's fidelity score.

flowchart LR
    A["Query & Response Segments"] --> B["Gemini API Call (url_context)"]
    C["Target URL"] --> B
    B --> D["JSON Output:<br/>• Survived<br/>• Lost<br/>• Distortions"]
    D --> E["Fidelity Score Calculation"]