<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Joachim Kolle</title>
    <description>The latest articles on DEV Community by Joachim Kolle (@joachim_k).</description>
    <link>https://dev.to/joachim_k</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3811468%2F6202e39c-4dc4-451c-b1d8-8cae70827265.jpeg</url>
      <title>DEV Community: Joachim Kolle</title>
      <link>https://dev.to/joachim_k</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/joachim_k"/>
    <language>en</language>
    <item>
      <title>The missing tool in my AI coding stack was Playwright</title>
      <dc:creator>Joachim Kolle</dc:creator>
      <pubDate>Sat, 07 Mar 2026 14:47:04 +0000</pubDate>
      <link>https://dev.to/joachim_k/the-missing-tool-in-my-ai-coding-stack-was-playwright-54ec</link>
      <guid>https://dev.to/joachim_k/the-missing-tool-in-my-ai-coding-stack-was-playwright-54ec</guid>
      <description>&lt;p&gt;AI is really good for frontend work. Definitely Opus 4.6. ChatGPT-5.4 is still catching up, but Claude has the crown right now, &lt;br&gt;
anyways.&lt;/p&gt;

&lt;p&gt;I use it a lot for design, UI ideas, and getting screens built fast.&lt;/p&gt;

&lt;p&gt;But the biggest problem shows up right after that.&lt;/p&gt;

&lt;p&gt;The UI looks done.&lt;br&gt;
The code looks fine.&lt;br&gt;
Nothing throws errors.&lt;/p&gt;

&lt;p&gt;And then the actual product is broken.&lt;/p&gt;

&lt;p&gt;I ran into this while building the application form in my open source ATS.&lt;/p&gt;

&lt;p&gt;Some fields just didn’t work.&lt;/p&gt;

&lt;p&gt;No big red error or obvious crash.&lt;/p&gt;

&lt;p&gt;The AI had written something that looked correct, but it never actually tested the flow.&lt;/p&gt;

&lt;p&gt;I used to just manually check after the AI has written the code.&lt;/p&gt;

&lt;p&gt;That’s basically the problem with AI coding right now.&lt;/p&gt;

&lt;p&gt;A lot of code is almost right.&lt;/p&gt;

&lt;p&gt;And i came accross a stack overflow survey found that 84% of developers use or plan to use AI tools, but more distrust the output than trust it: 46% vs 33%.&lt;/p&gt;

&lt;p&gt;What fixed this for me was Playwright.&lt;/p&gt;

&lt;p&gt;I started writing tests for the critical flows.&lt;/p&gt;

&lt;p&gt;For me that means stuff like:&lt;/p&gt;

&lt;p&gt;create a job with all custom fields -&amp;gt; apply to the job -&amp;gt; check the application in the dashboard&lt;/p&gt;

&lt;p&gt;Then I can be more confident when i make some changes, that the main features wroks.&lt;/p&gt;

&lt;p&gt;I write it once, and it keeps protecting the product on every commit.&lt;/p&gt;

&lt;p&gt;I run it in GitHub Actions on every pull request, which Playwright supports well in CI workflows.&lt;/p&gt;

&lt;p&gt;That changed the whole way I use AI.&lt;/p&gt;

&lt;p&gt;Now AI can help me move fast, but Playwright is the thing that tells me whether the feature actually works.&lt;/p&gt;

&lt;p&gt;It's also well integrated in vs code, and AI is also pretty good at writing the Playwright tests.&lt;/p&gt;

&lt;p&gt;You can basically one shot most tests, if there's any erros just write: playwright test it not working, fix. Make no mistakes!! (for the vibe coders ;)&lt;/p&gt;

&lt;p&gt;Hope this was valuabe :)&lt;/p&gt;

&lt;p&gt;Curious if anybody else have had this problem and how did you solve it?&lt;/p&gt;

</description>
      <category>webdev</category>
      <category>ai</category>
      <category>programming</category>
    </item>
  </channel>
</rss>
