Make ClawBench Braintrust publishing opt-in - #228
Closed
masnwilliams wants to merge 1 commit into
Closed
masnwilliams wants to merge 1 commit into
masnwilliams wants to merge 1 commit into
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Collaborator
Author
|
superseded, evals keep publishing |
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
The ClawBench workflow stops publishing every run to Braintrust. The publish step now runs only when the
BENCHMARK_PUBLISH_BRAINTRUSTrepository variable istrue.Why this is enough
The benchmark report is rendered from the Harbor job directories, not from Braintrust, so unpublished runs still get the same job summary and pull-request comment. The only thing they lose is the "Open the Braintrust experiment" link, which the report already omits when there is no publication file.
Changes
benchmark-clawbench.yml:if: vars.BENCHMARK_PUBLISH_BRAINTRUST == 'true'on Publish Braintrust experiment. Check benchmark execution now accepts askippedpublish as well as a successful one.benchmarks/harbor/README.md: documents the variable. Localbun run benchmark:publishis unchanged.Testing
actionlintpasses on the workflow. I have not triggered a benchmark run.Note
Low Risk
Workflow gating and documentation only; benchmark execution and reporting paths are unchanged when publishing is enabled.
Overview
ClawBench CI no longer publishes every benchmark run to Braintrust by default. The Publish Braintrust experiment step runs only when the repository variable
BENCHMARK_PUBLISH_BRAINTRUSTistrue; job summaries and PR comments still come from Harbor job directories, so skipping publish only drops the Braintrust experiment link when nopublication.jsonis produced.The final Check benchmark execution gate now treats a skipped publish step as acceptable (
successorskipped), so workflows can pass without Braintrust upload.benchmarks/harbor/README.mddocuments the opt-in variable for CI alongside the existing Braintrust credentials.Reviewed by Cursor Bugbot for commit 5cfd910. Bugbot is set up for automated code reviews on this repo. Configure here.