Test repair agent: Part1 - #6375
Merged
Merged
Conversation
marco-c
reviewed
Jul 24, 2026
marco-c
reviewed
Jul 24, 2026
marco-c
reviewed
Jul 24, 2026
marco-c
reviewed
Jul 24, 2026
marco-c
reviewed
Jul 28, 2026
marco-c
reviewed
Jul 28, 2026
marco-c
reviewed
Jul 28, 2026
marco-c
reviewed
Jul 28, 2026
Contributor
Author
|
I think some complexity in deterministic code can go away by using Treeherder CLI and/or API, as it provides some of the things we do manually out of the box. I've already switched to the API in the listener. I propose doing it as a follow-up, though, as a part of #6417 |
marco-c
approved these changes
Jul 29, 2026
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Analyzes test failure, identifies the culprit (including previous pushes), proposes a fix if it can and gives a recommendation what to do about it (back out, land fix, rerun etc.).
It follows the architecture of Build Repair Agent.
The most interesting parts are in "deterministic" code that prepares the data for the agent, as there might be bugs there. Still, we want to narrow it down for the agent to simplify it, make it less fragile, and spend fewer tokens.
The listener (Part 2) will filter non-regressions, intermittent failures etc., so that the API is triggered only for useful investigations.
I tested it on a few test failures locally, but we should deploy it and observe how it works with real-time failures in different environments.
Test runs in Weave (the last one with all the fixes)
#6399