Edition of Thursday 12 November 2026

Tests as Data

Comments on “Tests as Data”

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

Matteo Vaccari, writing on martinfowler.com, argues that AI-written tests are hard to trust because they are voluminous, hard to read, coupled to implementation and often just confirm current behavior. He proposes customer-style black-box tests kept as data (tables, YAML, or small text DSLs) run by a short runner an agent can generate, making changes easier to spot in review. His support is interviews with 26 Thoughtworks developers and worked examples; the quality-decline reports he cites are not sourced, and he offers no measured results.

Read the original at martinfowler.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. 16 of 106 with a score of 0.69. The panel rated it highest on new information (0.88) and practical relevance (0.81), lowest on substantiated (0.49). Placed No. 16, No. 14 by score alone: the fifth AI & machine learning story of the day (−0.040). The placement terms move stories; they are not part of the score.

AI judgeJudge 2claude-haiku-5-5
Practitioner essay on making AI-generated tests trustworthy via tests-as-data, customer tests and DSLs; well-referenced and actionable, though mostly argument and framing rather than new measured evidence.
AI judgeJudge 3gpt-5.6-luna
A detailed, actionable engineering argument for tests-as-data and executable customer specifications in AI-assisted development, with concrete DSL examples and testing practices. It is opinionated and partly anecdotal, but technically rich and relevant.

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

0.69total under rubric v2.3

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.60
  • Reader interest0.78
  • Audience breadth0.72
  • Practical relevance0.81
  • Substantiated0.49
  • Technical depth0.76
  • New information0.88
  • Counts against: Promotional0.05
  • Counts against: Overclaim0.28

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?0.650.450.720.600.27+0.24+0.145
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.690.800.850.780.16+0.26+0.203
Audience breadthHow many people who build technology does it concern?0.840.700.620.720.22+0.13+0.094
Practical relevanceHow useful is it to someone who builds software or tech products?0.860.800.780.810.08+0.22+0.179
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.0.360.450.670.490.31+0.05+0.025
Technical depthHow technically substantive is it?0.830.650.800.760.18+0.05+0.038
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.880.850.900.880.05+0.05+0.044
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.0.080.050.030.050.05−0.19−0.010
OverclaimDo the headline or conclusions claim more than the body supports?0.280.200.350.280.15−0.11−0.030
Sum of contributions+0.687
Claim-check penalty0.000
Total, as published0.687

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 #16 on Rubric, 12 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-12/16/"><img src="https://rubric.hrih.si/badge/2026-11-12/16.svg" alt="Ranked #16 on Rubric, 12 Nov 2026" width="245" height="28"></a>

Claim check

Claim check: 4 no source given · 4 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.

  1. No source givenNumber

    In early 2026 the author interviewed 26 developers at an organization heavily invested in AI coding agents, and they invariably said they did not trust their tests much.

    The article gives no link or citation for it.

    The article says

    “In early 2026, I conducted interviews with 26 developers working for an organization that was heavily invested in developing with AI coding agents”

    The source says

    No source to compare with.

  2. No source givenGeneralization

    Several reports published in early 2026 show that the quality of code being deployed by organizations is getting worse.

    The article gives no link or citation for it.

    The article says

    “there have been a number of reports published in early 2026 showing that the quality of code being deployed by organizations is getting worse”

    The source says

    No source to compare with.

  3. No source givenCausal conclusion

    Test-first and test-driven development allowed teams building business software to deliver nearly bug-free code at a fraction of the cost of formal specifications and proofs.

    The article gives no link or citation for it.

    The article says

    “This practice allowed teams developing business software to deliver nearly bug-free code, at a fraction of the cost of formal specifications and proofs”

    The source says

    No source to compare with.

  4. Couldn’t checkCausal conclusion

    Coding agents perform better when told what is wanted and how to validate it, while adding instructions on how to achieve the result seems to make them perform worse.

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

    The article says

    “adding instructions on how to achieve the result seems to make them perform worse”

    The source says

    No passage could be compared.

    Source cited by the article: danluu.com/agentic-testing

  5. No source givenDate

    Around 2002 Ward Cunningham wrote a program that read test data from Excel and executed tests automatically, and called the tool Fit.

    The article gives no link or citation for it.

    The article says

    “Cunningham tells about how around the year 2002, his customers tended to provide their own specifications in Excel documents”

    The source says

    No source to compare with.

  6. Couldn’t checkDate

    In 2004 Martin Fowler published his article on Specification by Example.

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

    The article says

    “In 2004, Martin Fowler published his article on Specification by Example”

    The source says

    No passage could be compared.

    Source cited by the article: martinfowler.com/bliki/SpecificationByExample.html

  7. Couldn’t checkQuote

    James Shore became disillusioned with Fit because most customers did not want to write the test tables, so he wrote the Fit tests himself.

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

    The article says

    “most customers did not want to write the test tables, which led him to have the conversations with the customers and then write the Fit tests himself”

    The source says

    No passage could be compared.

    Source cited by the article: jamesshore.com/v2/blog/2010/the-problems-with-acceptance-testing

  8. Couldn’t checkVersion

    Browser-based component testing was recently promoted to Trial in the Thoughtworks Technology Radar.

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

    The article says

    “Browser-based component testing was recently promoted to Trial in the Thoughtworks Technology Radar”

    The source says

    No passage could be compared.

    Source cited by the article: thoughtworks.com/radar/techniques/browser-based-component-testing

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN 2 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 12 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 “Tests as Data”

More on Rubric

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

  1. Score 0.75

    llama.cpp v1.0