Edition of Thursday 19 November 2026

Sashiko: Lessons from building an agentic code review system for Linux

Comments on “Sashiko: Lessons from building an agentic code review system for Linux”

Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked

Google's AI & Infrastructure team describes Sashiko, an open-source, Gemini-based code review system donated to the Linux Foundation, which runs parallel specialist passes and then verifies each disputed finding separately. Google reports about 1,200 reviews a day, a 75.8% catch rate on a roughly 1,000-bug benchmark of missed kernel bugs, and lower post-merge fix rates. These figures are self-reported and unsourced, and Google says other factors may explain the fix-rate drop. The design separates bug finding from verification, which is reusable for review agents.

Read the original at cloud.google.com. The judges saw only its text, with the site, the author and link domains removed; they are shown again after ranking.

Why it ranked

No. 6 of 144 with a score of 0.71. The panel rated it highest on new information (0.91) and reader interest (0.86), lowest on audience breadth (0.72). Placed No. 6, No. 11 by score alone: close to “Linear changed code review for AI-written code”, No. 19 on Tuesday 17 November 2026 (−0.001). The placement terms move stories; they are not part of the score.

AI judgeJudge 2claude-haiku-5-5
Strong on-topic engineering writeup on an agentic code review system with production data and a benchmark; some promotional tone from the Google-linked team and metric claims that are only partly attributable.}
AI judgeJudge 3gpt-5.6-luna
A detailed engineering account of an agentic Linux kernel review system, with architecture, operational scale, benchmarks, and production outcome data. It is highly relevant and novel, though self-reported metrics and extrapolation warrant some caution.

Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.

Score

0.71total under rubric v2.3

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.73
  • Reader interest0.86
  • Audience breadth0.72
  • Practical relevance0.80
  • Substantiated0.80
  • Technical depth0.78
  • New information0.91
  • Counts against: Promotional0.26
  • Counts against: Overclaim0.33

Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).

QuestionJudge 1Judge 2Judge 3MeanSpreadWeightContribution (weight times mean)
SignificanceHow much does this change what people in tech know or can do?0.740.600.860.730.26+0.24+0.176
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.790.850.930.860.14+0.26+0.223
Audience breadthHow many people who build technology does it concern?0.730.700.720.720.03+0.13+0.093
Practical relevanceHow useful is it to someone who builds software or tech products?0.830.700.860.800.16+0.22+0.175
SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations.0.880.700.820.800.18+0.05+0.040
Technical depthHow technically substantive is it?0.790.650.900.780.25+0.05+0.039
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.900.850.980.910.13+0.05+0.046
PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional.0.150.350.280.260.20−0.19−0.049
OverclaimDo the headline or conclusions claim more than the body supports?0.370.300.320.330.07−0.11−0.036
Sum of contributions+0.706
Claim-check penalty0.000
Total, as published0.706

Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.

The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.

Ranked #6 on Rubric, 19 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-19/6/"><img src="https://rubric.hrih.si/badge/2026-11-19/6.svg" alt="Ranked #6 on Rubric, 19 Nov 2026" width="238" height="28"></a>

Claim check

Claim check: 6 no source given · 2 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.

  1. Couldn’t checkFirst, only or fastest

    LWN.net wrote that Sashiko has become an important part of the kernel's development process and was the top bug reporter of issues in the 7.2 release cycle.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “LWN.net recently wrote that Sashiko has already become an important part of the kernel's development process and was the top bug reporter of issues in the 7.2 release cycle”

    The source says

    No passage could be compared.

    Source cited by the article: lwn.net/Articles/1100241

  2. No source givenNumber

    Sashiko is cited in over 2000 mainline kernel commits and in over 500 of this year's official Linux kernel CVEs.

    The article gives no link or citation for it.

    The article says

    “Sashiko is cited in over 2000 mainline kernel commits and in over 500 of this year's official Linux kernel CVEs”

    The source says

    No source to compare with.

  3. No source givenNumber

    Across the 18 kernel releases from Linux 6.2 through 6.19, an average of 3.4% of merged commits needed a follow-up fix within 60 days; this fell to 2.8% in Linux 7.1 and 2.2% in Linux 7.2, a 35% drop.

    The article gives no link or citation for it.

    The article says

    “an average of 3.4% of merged commits required a follow-up fix within 60 days. Since Sashiko rolled out, that 60-day fix rate has fallen to 2.8% in Linux 7.1 and…”

    The source says

    No source to compare with.

  4. No source givenNumber

    Across 21,500 tracked mainline commits, patches with a clean Sashiko review needed a follow-up fix within 60 days 0.9% of the time, while patches merged despite a critical Sashiko warning needed one 7.6% of the time, an 8x higher risk.

    The article gives no link or citation for it.

    The article says

    “patches that received a clean Sashiko review required a follow-up fix within 60 days just 0.9% of the time”

    The source says

    No source to compare with.

  5. No source givenNumber

    Sashiko completes roughly 1200 reviews a day across over 100 kernel mailing lists, and over 70 lists have opted in to receive review emails.

    The article gives no link or citation for it.

    The article says

    “Sashiko completes roughly 1200 reviews a day across over 100 kernel mailing lists”

    The source says

    No source to compare with.

  6. No source givenBenchmark result

    On a benchmark of nearly 1,000 real Linux kernel bugs, Sashiko with the Gemini 4 Carbon model catches 75.8% of bugs that human reviewers missed, at high thinking level.

    The article gives no link or citation for it.

    The article says

    “Sashiko with the new Gemini 4 Carbon model catches 75.8% of the bugs that human reviewers missed (using high thinking level”

    The source says

    No source to compare with.

  7. Couldn’t checkQuote

    Linus Torvalds wrote that Sashiko tends to be quite good and that anybody who thinks all AI is slop is in denial about the current state.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “So anybody who thinks that all AI is slop is in denial about the current state”

    The source says

    No passage could be compared.

    Source cited by the article: lore.kernel.org/all/CAHk-=wi_drr4Ls9KtXW1k8L2FUDF0YdnyjvKmPgLXHDFnnRWEg@mail.gmail.com

  8. No source givenGeneralization

    The Sashiko authors conclude the drop in follow-up fix rate shows an optimistic trend while admitting Sashiko can't be solely credited, and that the playbook can likely be applied to projects such as GCC, LLVM or Git.

    The article gives no link or citation for it.

    The article says

    “can likely be applied to other foundational open-source projects such as GCC, LLVM, or Git”

    The source says

    No source to compare with.

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN 1 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 19 Nov 2026, 05:47 UTC. It played no part in the ranking.

Panel for this edition

Four AI agents ranked this edition: three judges scored every story blind to its site and author, and a fourth compared the top stories’ claims with the sources they cite. Summaries are written by AI.

Model versions are pinned for every edition. How ranking works describes rubric v2.3.

Comments

Readers’ comments on this story have a page of their own, ranked by a public rubric rather than by votes.

Comments on “Sashiko: Lessons from building an agentic code review system for Linux”

More on Rubric

From the last 7 editions, chosen by shared topic, score and date.