Try the Tech Stack Dataset Item API

- The item carries `domain`, `tld`, `rank`, `is_infrastructure` (`true` when the domain looks like backend CDN/DNS/cloud-vendor infrastructure — e.g. `cloudfront.net`, `gtld-servers.net` — rather than a real, human-navigable site; a best-effort heuristic classification, so `false`/absent is not a guarantee the domain is a real site), `reachable`, `status` (HTTP status of the last probe attempt, when one was made), `technologies` (each with `name`, `categories`, `confidence`, `version`, `evidence`), `technology_names`, `categories`, `tech_count`, the `cms`/`ecommerce`/`cdn`/`web_server`/`server_language` rollups, `render_tier`, `detector_version`, and the crawl provenance (`run_id`, `crawled_at`). - When the last probe attempt didn't fully pass, the item also carries `failure_reason` (a short classification, e.g. `"blocked/challenge markers found"`, `"bad status 503"`, `"request failed"`), `probe_error` (the first line of the underlying error, when there was one), and `block_matched` (which block/challenge marker(s) were found, when `failure_reason` is a block/challenge classification). This can appear even when `reachable` is `true` — a blocked page returning a non-5xx status still counts as reachable while carrying a failure reason. Absent for a site that fully passed. - When the crawl looked for extra non-homepage pages (opt-in per run), the item also carries `secondary_pages_discovered` (count of candidate pages found on the homepage nav, regardless of fetch outcome), `secondary_pages_fetched` (count that were actually fetched successfully and contributed content — always <= discovered), `secondary_page_categories` (which categories were successfully fetched and contributed, e.g. `login`, `docs`, `cart`), `secondary_pages` (the full per-page audit trail — every discovered candidate, each entry `{category, url, fetched}`; `fetched: false` means the candidate was found on the homepage nav but the fetch itself failed, e.g. blocked or timed out), and `secondary_only_technology_names` (technology names detected only via a fetched secondary page, not the homepage itself — a subset of `technology_names`). Absent when no secondary pages were discovered for that run. - Returns the standard not-found envelope when the domain is not in the index. - Does not trigger live scraping; for a live, on-demand fingerprint of any URL use `/web/techstack`.