Skip to content

Add OpenAI TTL cache metric specifications - #89

Draft
Andrew Kent (realark) wants to merge 1 commit into
mainfrom
ark/sdk-844-openai-cache-metrics
Draft

Andrew Kent (realark) wants to merge 1 commit into
mainfrom
ark/sdk-844-openai-cache-metrics

Conversation

@realark

@realark Andrew Kent (realark) commented Nov 17, 2026 •

Copy link
Copy Markdown
Member

AI Generated Description

Summary

  • Define OpenAI 30m/12h cache-write metric mappings, reported-zero handling, aggregate fallback, and inclusive prompt-token accounting.
  • Add BTX scenarios for non-streaming 30m writes and streaming 12h writes.

Verification

  • Recorded real OpenAI Responses API responses using the Java SDK.
  • Both new BTX scenarios pass in replay mode, including Java fetching this published spec revision.

Dependencies

- Define 30m/12h usage mapping and single-representation token accounting.
- Add non-streaming 30m and streaming 12h BTX scenarios verified with Java.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant