/**
 * PDF extraction — discovers and extracts text from PDF URLs found during crawl.
 *
 * When the HTTP engine or browser engine discovers a link to a .pdf file,
 * this module fetches the PDF, extracts its text content, and returns it
 * as a screen-compatible markdown string for test generation.
 *
 * Uses pdf-parse (already installed in API package, accessible via dynamic import).
 */
export interface PdfScreen {
    url: string;
    title: string;
    markdown: string;
    pageCount: number;
    info: Record<string, string>;
    sizeBytes: number;
}
export declare function isPdfUrl(url: string, contentType?: string): boolean;
export declare function extractPdfFromUrl(url: string, opts?: {
    userAgent?: string;
    timeoutMs?: number;
    maxSizeBytes?: number;
}): Promise<PdfScreen | null>;
/**
 * Filter a list of discovered URLs to find PDF links.
 * Used by http-engine and browser crawler to decide which links to pass to extractPdfFromUrl.
 */
export declare function filterPdfUrls(urls: string[]): string[];