Repository Issue Activity (beta)

#3867 [Pytorch] There are 20 Recurrently Failing Jobs on pytorch/pytorch nightly

#6716 [Pytorch] There are 2 machines with long queues

#4179 [Pytorch] There are 2 Recurrently Failing Jobs on pytorch/pytorch main

#8560 Preparing a docs on how to improve macOS queue time and TTS

#8407 Investigation: Flaky test bot logic and instrumentation

#8296 Code-side decorator for labeling tests to be run periodically

#8281 Formalize workflow/job reliability standard metrics for QoS/flakiness tagging

#8590 Update /reliability page to also track workflows marked as UNSTABLE

#8455 Add AI alerts for key tracked metrics

#8570 Update pr-review skill to account for test best practices

#8586 Greenlight - Timeout on review jobs

#8534 I don't think we should re-approve an already approve PR to reduce spam. Also we most likely don't want the bot to un-approve/re-approve in a loop https://github.com/pytorch/pytorch/pull/193618#pullrequestreview-4940815976

#8524 Review after pushed changes

#8581 Greenlight - a merged/reverted PR should not be allowed to be greenlight stampped

#8573 Greenlight - use DrCI comment to update status

#8572 Greenlight stop double-accepting PRs

#8433 [Greenlight] Security hardening on agent flow

#8571 ai framework: dispatcher PoC

#8351 ai framework: evaluate Vercel sandboxes

#8563 Treehugger agent to monitor and alert MacOS spa workflow runs

#8566 Treehugger agent to monitor MacOS spa and health

#8516 Update HUD metrics board queueing table to include adjacent repos like executorch

#8561 Backfill + update lambda to keep flakiness data for hud using AI Advisor

#8562 Automatically merge PR if Dr. CI AI verdict deems breaking signals are unrelated

#8458 Autorevert data on hud to help distinguish between infra and test flakiness on hud