Sashiko that was built by Google engineers for agentic review of the Linux kernel code changes and powered by Google's AI models continues proving very valuable to kernel developers. In less than one year since they launched Sashiko for agentic AI code review, their results are quite fascinating.

Google engineer Roman Gushchin provided a status update on Sashiko at this week's Linux Plumbers Conference. This open-source code that works with Google's Gemini and other LLMs has completed more than 191k patch reviews across 99 different mailing lists in the less than one year it's been operating. As part of the review process it's performed more than 19 million autonomous Git lookups.

Sashiko has also received more than 7.6k replies from more than 1.1k kernel developers in response to Sashiko reports.

This year there have been more than 1,277 commits in the upstream Linux kernel that cite Sashiko. Additionally, 1,567 commits in linux-next citing it.

There have also been 463 CVEs so far this year to the Linux kernel that cite Sashiko.

In moving forward, Google engineers are working on a local/terminal review mode for Sashiko as well as a persistent bug database and even dog-feeding Sashiko for Sashiko as a self-review process.

Those wanting to learn more about Sashiko as we approach the end of 2026 can find all the details via the LPC 2026 presentation material.