FFlyType
macOSWindowsiOSAndroidHub

Site analyze and selective ingest

Discover pages across a whole site, preview titles and quality hints, then ingest only what you need — no live mirror of the site.

Before you start

  • Signed in with a Mind notebook open.
  • Root URL must be https (or http) with same-origin links only.

Steps

  1. 1

    Mind → notebook → Sources (Hub) or Add sources → Websites (desktop/mobile).

  2. 2

    Enter the site root URL (e.g. https://docs.example.com) → Analyze site.

  3. 3

    Server reads robots.txt, sitemap.xml when present, then BFS same-origin crawl up to plan limit.

  4. 4

    Preview lists title, page type, quality hint — toggle pages on or off.

  5. 5

    Ingest selected sends only chosen URLs; each becomes a normal web source (skip_analyze on server).

  6. 6

    Single Add website still ingests one URL immediately without the analyze step.

  7. 7

    Failed URLs show in source status — Reprocess from the source row.

Common mistakes

  • Trial discovers up to ~25 pages and ingests ~10 per batch; Pro allows larger discovery and ingest.
  • Prefer analyze for docs sites; use site-pack paste for a hand-picked URL list.
  • Chat citations use indexed text — we do not claim a live mirror of the site.

Related articles