DoiT consoleOpenAPISupport
Changelog
About
Start typing to search…

Ephemeral conversations for Ask Ava

Updated the Ask Ava endpoint to better support programmatic consumers like the Terraform provider, MCP server, and CI pipelines.

The sync endpoint created conversations that some automated consumers could not clean up, leading to orphaned conversations in users’ history. The new ephemeral flag lets these callers skip persistence entirely, while the conversationId field enables explicit cleanup for non-ephemeral use cases.

  • Added ephemeral: boolean request flag (default false).

    • When set to true: Conversations are not persisted. Recommended for programmatic consumers (Terraform, MCP, CI pipelines).
    • When set to false: Conversations are persisted and the response includes conversationId so callers can manage or delete conversations.

  • Changed the response body from a plain string to:

    { "answer": "string", "conversationId": "string (optional)" }


Support operations

New Support endpoint operations:

New schemas:

  • TicketDetailExtAPI — Extend TicketListItem with description.
  • CommentExtAPI — Comment with author email and Unix ms timestamp.
  • CommentsResponseExtAPI — Comments array (no pagination).
  • CreateCommentRequest — Body and optional private flag.

Dimension types

Added:

  • allocation_rule and allocation as aliases for the deprecated (but supported) attribution and attribution_group dimension types, to align the Cloud Analytics APIs with Allocations.
  • stricter validation for dimension lookups (GET /analytics/v1/dimension), the type and ID must refer to a real dimension pair. 400 - Bad Request is returned for invalid combinations.

These changes affect the Dimensions, Reports, Alerts, and Budgets APIs wherever dimension or filter types are used.


Pagination max results

For pagination the default maxResults is now set to 50, instead of 500. maxCap is still 500.