Skip to content

GraphQL API for AI requests

Jan Provaznik requested to merge jp-ai-cache-api into master

What does this MR do and why?

Related to https://gitlab.com/gitlab-org/gitlab/-/issues/410521

Adds API for accessing cached AI requests/responses.

How to set up and validate locally

  1. Enable caching of AI requests 
    Feature.enable(:ai_redis_cache)
  2. Trigger an AI action (for example "summarize comments" on issue show page)
  3. Check that request and response are returned with following query:
query {
  aiMessages {
    nodes {
      id
      requestId
      content
      role
      errors
    }
  }
}

Or get only bot response for a specific request id:

query {
  aiMessages(requestIds: ["999dac16-0893-4d11-9e01-2ab2a3c9128b"], roles: [BOT]) {
    nodes {
      id
      requestId
      content
      role
      errors
    }
  }
}

MR acceptance checklist

This checklist encourages us to confirm any changes have been analyzed to reduce risks in quality, performance, reliability, security, and maintainability.

Edited by Jan Provaznik

Merge request reports