nx_revimg_crawl_run.nx

nx_revimg_crawl_run.nx -- LIVE reverse-image discovery CLI: the operator-driven assembly that runs the whole loop on the real web. Given a SEED page (the entity's media, e.g. the Diora Baird page) and one or more CANDIDATE site pages, it: fetches each over the team's OWN validated HTTPS (sovereign DNS+TCP+ TLS1.3+X.509, the nx_media_gather path), extracts image URLs (nx_html_extract_imgs), politely fetches each image, decodes it (nx_img_to_gray -- PNG today, JPEG since the FF-D8-FF wiring, so the real web works), perceptual-hashes it (nx_phash dHash), then runs nx_revimg_discover to report WHICH candidate sites host the seed media (recompression/format-robust, which a sha256 byte-hash can't) and HARVESTS those hosts' OTHER media -- the "grow the search results" answer. SAFETY: with no args it prints usage and exits (the build-time self-run never touches the network). BOUNDED: <= MAX_PER_PAGE images per page, polite 1.2s pacing, total fetches kept under the per-process certloop ceiling (~19) -- the cap is LOGGED, never silent. Composes nx_media_gather's HTTPS pattern + nx_revimg_crawl/discover -- ZERO new transport or ranking logic. Usage: nx_revimg_crawl_run <seed_page_url> [candidate_url ...] license_tier: ORIGINAL