Skip to main content
Every GET endpoint that returns a collection paginates with opaque cursors. There are no page numbers, no offsets, no totals — cursors are forward-only pointers into the result set.

Parameters

Limit validation

Anton currently has two limit-handling paths across list endpoints. Both cap the page at 100 items, but they differ on how an overflow is surfaced: Guidance for integrators: treat the strict contract as authoritative. Always request limit values in the 1..100 range. Do not rely on the silent-cap behavior — Anton may migrate all endpoints to strict validation in a future release, at which point over-limit requests will uniformly return 400 invalid_limit.

Response envelope

The last page omits next_cursor and sets has_more to false:

Iterating through pages

Rules

Set limit=100 to minimize round trips when fetching large result sets.
has_more is the only correct termination signal. Don’t compare len(data) to limit — the last page may be full.
Treat cursor values as opaque strings. Do not parse, modify, or construct them. Always pass the exact next_cursor value returned by the API.
Cursor values remain valid indefinitely. You can store a cursor and resume pagination later — useful for long-running exports.