<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <id>https://stackassay.com/stories/</id>
  <title>Stack Assay: Lab stories</title>
  <subtitle>Why an AI trend matters, who the players are, what the outside evidence says and what our lab measured, one trend at a time.</subtitle>
  <link rel="self" href="https://stackassay.com/stories/feed.xml"/>
  <link rel="alternate" type="text/html" href="https://stackassay.com/stories/"/>
  <author><name>Stack Assay</name></author>
  <rights>Text under CC BY 4.0</rights>
  <updated>2026-11-20T00:00:00Z</updated>
  <entry>
    <id>https://stackassay.com/stories/harness-bench-to-stack/</id>
    <title>The harness is the new product layer. We measured four of them, then built one.</title>
    <link rel="alternate" type="text/html" href="https://stackassay.com/stories/harness-bench-to-stack/"/>
    <link rel="alternate" type="text/markdown" href="https://stackassay.com/stories/harness-bench-to-stack.md"/>
    <published>2026-11-20T00:00:00Z</published>
    <updated>2026-11-20T00:00:00Z</updated>
    <summary>In 2026 most professional developers use a coding agent every week, and the product they choose is less a model than a harness: the loop, tools, memory and sandbox around the model. Vendors, open-source projects and gateways now compete on that layer, and public benchmarks disagree on how much it matters. We studied the trend, then ran four leading harnesses on the same model. All four solved all 168 of our runs; they differed in cost, by 4.9× per solved task, mostly through prompt caching. An independent paper published weeks earlier found the same shape across seven models. We packaged what we learned as the agent harnesses Stack and measured it in public beside the others.</summary>
  </entry>
</feed>
