dabla opened a new pull request, #74579: URL: https://github.com/apache/airflow/pull/74579
First step toward #74562: until the micro-benchmark suite proposed there lands, this asks agents to measure performance-sensitive paths themselves, so AI-assisted changes do not skip it. Suggested in the contributors Slack. The pre-push review of the `airflow-publish-changes` skill gets a new step and a "Performance-sensitive components" section: - the components that run for every task, Task SDK call, log line, Dag parse or scheduler loop (Task SDK execution time, Execution API, logging and secret masking, serialization, scheduler and triggerer loops, Dag processing), with their paths; - how to measure: the same script on the base commit and on the change, alternating runs, medians, a control the change does not touch, `timeit` or the existing `perf_kit` helpers, `assert_queries_count` for database code; - a before/after table in the PR description, with the reason when the change is slower. Only the skill changes, not `AGENTS.md`, so the skill-eval hash stays as it is. Once #74562 is in, the section can point at the `benchmarks/` folders instead. --- ##### Was generative AI tooling used to co-author this PR? - [X] Yes — Claude Code (Opus 5.5) Generated-by: Claude Code (Opus 5.5) following [the guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
