Banned domains are detected by scanning the full generated draft against the banned co-link list using word-boundary regular expressions that catch base domains and subdomains without falsely matching partial brand names, immediately folding the draft if any match is found.

flowchart TD
    A[Draft Generation] --> B[Regex Scan with Word Boundaries]
    B --> C{Banned Domain Match?}
    C -- Yes --> D[Fold Draft: banned-link Verdict]
    C -- No --> E[Proceed to Blue Team Evaluation]