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)#
Endpoint:
base_url+list_path(e.g.https://api.example.com+/v1/tickets; a fullhttp…URL inlist_pathis used as-is). Reads withGETby default,POSTif the API wants a body.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.Find the items:
items_pathpoints at the array (/data,/issues). Leave it empty and the connector tries the body itself, then/data,/items,/results.Map the fields with JSON pointers (
/a/bverbatim,a.bshorthand):Setting Default Notes map_id/idWhat identifies an item; re-syncs update instead of duplicating map_title/titlemap_text/bodyComma-separated pointers are joined, e.g. description, commentsmap_author/authormap_updated/updated_atLast-changed date; drives incremental sync 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.