Guides Connect any JSON API to AnchorCore

Any JSON API → AnchorCore

Point AnchorCore at a JSON list API — tickets, issues, headless CMS, your own backend. Map its fields once, preview the result, sync. No per-integration code.

Setup (Sources → Add source → Generic REST API)#

  1. Endpoint: base_url + list_path (e.g. https://api.example.com + /v1/tickets; a full http… URL in list_path is used as-is). Reads with GET by default, POST if the API wants a body.

  2. Auth: none, Bearer token, Basic (email + password), or a custom header (default X-Api-Key). Tokens and passwords are kept as secrets, never in the database: in your OS keychain on macOS and Windows, elsewhere in an encrypted file in your data folder.

  3. Find the items: items_path points at the array (/data, /issues). Leave it empty and the connector tries the body itself, then /data, /items, /results.

  4. Map the fields with JSON pointers (/a/b verbatim, a.b shorthand):

    SettingDefaultNotes
    map_id/idWhat identifies an item; re-syncs update instead of duplicating
    map_title/title
    map_text/bodyComma-separated pointers are joined, e.g. description, comments
    map_author/author
    map_updated/updated_atLast-changed date; drives incremental sync
  5. Preview, then Sync. Preview fetches one page and shows the first 5 mapped records without saving anything. When the mapping looks right, save and Sync — items appear in answers cited as rest:<id>.

Pagination & incremental sync#

page_mode: none, token, page, or offset — with all parameter names overridable and max_pages bounding the loop (default 20, up to 200). since_param sends the sync cursor back to the API for incremental pulls.

What you get downstream#

REST items flow through the same pipeline as every other source: read, understood, cited as rest:<id> in answers, and gated by the same source labels and privacy rules. Items with no title and no text are skipped.

Source: docs/guides/rest-api.md in the public repo.