Skip to main content
The Attio source reads CRM objects and records, lists, notes, tasks, meetings, and call transcripts. It is built with an HTTP manifest and is alpha, so its resources and columns may change between releases.

​
Configuration

​
Resources

Some selections produce many API calls. objects feeds records, which feed record_entries, files, and threads; meetings similarly feed recordings and transcripts. Selecting the three per-record children for 50,000 records requires at least 150,000 child requests. Select only the resources you need.

​
Modes

Full only. The manifest declares no incremental blocks, so every run re-reads every selected resource from the start. Full runs checkpoint their pagination cursor per top-level resource and resume from the last completed page. Child resources always restart with their parent.

​
Behavior

  • Auth: bearer token from api_key on every request.
  • Pagination: offset paging on the query endpoints (injected into the POST body for records and entries, into the query string for attribute and entry lists), cursor paging with null termination on views, files, meetings, and call recordings, and single-page responses elsewhere.
  • Rate limiting: the manifest configures no client-side rate limit. The engine still backs off on 429 responses, honoring Retry-After for up to 10 attempts, but deeply nested resources still send requests as fast as Attio accepts them. Large workspaces may be throttled frequently.
  • Every resource carries a raw remainder column preserving fields the manifest does not project.