Anthropic connector FAQ
This feature is in Beta. Workspace admins can control access to this feature from the Previews page. See Manage Databricks previews.
This page answers frequently asked questions about the managed Anthropic ingestion connector in Lakeflow Connect. For FAQs that apply to all managed connectors, see Managed connector FAQs.
What tables does the connector support?
The connector supports the following Anthropic tables:
audit_log_activitiesorganizationsusersrolesgroupssettingschatsmessages
For the schema of each table, see Supported source tables.
Does the connector support incremental ingestion?
Yes, for a subset of tables. The audit_log_activities, chats, and messages tables are ingested incrementally using a cursor field (created_at, updated_at, and created_at, respectively). The directory tables (organizations, users, roles, groups, and settings) are refreshed with a full snapshot on each pipeline update.
What Anthropic access does the connector require?
The connector requires an Anthropic organization with Compliance API access enabled. If your organization doesn't have Compliance API access, see Anthropic: Compliance API access.
What Anthropic permissions do I need?
To create the Compliance Access Key that the connector uses, you must be an Anthropic organization admin. The key must be a Compliance Access Key (sk-ant-api01-...) created in claude.ai; Admin API keys (sk-ant-admin01-...) are not accepted. Grant the read:compliance_org_data scope for organization directory data and the read:compliance_user_data scope for user data. See Configure Anthropic.