Skip to content

trace scoring#90

Open
Andrew Kent (realark) wants to merge 1 commit intomainfrom
ark/trace-scoring
Open

trace scoring#90
Andrew Kent (realark) wants to merge 1 commit intomainfrom
ark/trace-scoring

Conversation

@realark
Copy link
Copy Markdown
Collaborator

No description provided.

@realark Andrew Kent (realark) added the enhancement New feature or request label Apr 23, 2026
@realark Andrew Kent (realark) force-pushed the ark/trace-scoring branch 12 times, most recently from f1fbd7c to c49a985 Compare April 28, 2026 20:15
@realark Andrew Kent (realark) changed the base branch from main to ark/api-client-codegen April 28, 2026 20:15
@realark Andrew Kent (realark) force-pushed the ark/trace-scoring branch 9 times, most recently from ee4746f to 68f6ec1 Compare April 30, 2026 06:49
Base automatically changed from ark/api-client-codegen to main April 30, 2026 18:26
@realark Andrew Kent (realark) force-pushed the ark/trace-scoring branch 4 times, most recently from 5b7070c to 8e0dd94 Compare May 1, 2026 21:26
@realark Andrew Kent (realark) marked this pull request as ready for review May 1, 2026 22:00
int lastInputLength = 0;

for (var span : sorted) {
// Extract new input messages (only those beyond what prior spans already contributed)
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

won't this break if there are independent llm spans happening? Like with subagents? Or am I thinking about this incorrectly?

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah that would break. I updated the logic to do a proper tree traversal + dedup. Did another manual test against an opencode trace with subagents: https://www.braintrust.dev/app/braintrustdata.com/p/andrew-misc/logs?tvt=trace&r=e6ca417d-e5e4-45b8-8e5f-90fb1e4f703e&s=e6ca417d-e5e4-45b8-8e5f-90fb1e4f703e

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants