<?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: Eduardo Dos Santos</title>
    <description>The latest articles on DEV Community by Eduardo Dos Santos (@eduardo_dossantos_0fc44c).</description>
    <link>https://dev.to/eduardo_dossantos_0fc44c</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.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3579185%2F0cb9298e-a0de-4ed2-a7f2-96fe23ba9ad7.jpg</url>
      <title>DEV Community: Eduardo Dos Santos</title>
      <link>https://dev.to/eduardo_dossantos_0fc44c</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/eduardo_dossantos_0fc44c"/>
    <language>en</language>
    <item>
      <title>I built PROOF: a tool to verify AI-generated apps with real browser tests</title>
      <dc:creator>Eduardo Dos Santos</dc:creator>
      <pubDate>Tue, 17 Nov 2026 15:04:32 +0000</pubDate>
      <link>https://dev.to/eduardo_dossantos_0fc44c/i-built-proof-a-tool-to-verify-ai-generated-apps-with-real-browser-tests-452o</link>
      <guid>https://dev.to/eduardo_dossantos_0fc44c/i-built-proof-a-tool-to-verify-ai-generated-apps-with-real-browser-tests-452o</guid>
      <description>&lt;p&gt;Building software with AI coding agents is incredibly fast.&lt;/p&gt;

&lt;p&gt;But I kept running into a frustrating problem:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The AI says the task is complete. The code compiles. The tests pass. But does the application actually work?&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;That question inspired me to start building &lt;strong&gt;PROOF&lt;/strong&gt;, a tool that turns real browser tests into visual evidence.&lt;/p&gt;

&lt;h2&gt;
  
  
  🔍 What is PROOF?
&lt;/h2&gt;

&lt;p&gt;PROOF is a local web application that uses Playwright to verify user interactions and generate reports showing exactly what was tested.&lt;/p&gt;

&lt;p&gt;Instead of relying only on statements like "all tests passed," I wanted a way to see the actual results.&lt;/p&gt;

&lt;p&gt;The current prototype can:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Execute configurable browser tests.&lt;/li&gt;
&lt;li&gt;Interact with real web applications.&lt;/li&gt;
&lt;li&gt;Verify specific success conditions.&lt;/li&gt;
&lt;li&gt;Capture screenshots during execution.&lt;/li&gt;
&lt;li&gt;Identify failed verification steps.&lt;/li&gt;
&lt;li&gt;Generate standalone HTML reports.&lt;/li&gt;
&lt;li&gt;Preserve evidence from completed test sessions.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  🧪 Testing a real application
&lt;/h2&gt;

&lt;p&gt;For the first independent test, I used TodoMVC, an open-source example application.&lt;/p&gt;

&lt;p&gt;I configured PROOF to:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Open the application in Chromium.&lt;/li&gt;
&lt;li&gt;Create a task called "Review the release checklist."&lt;/li&gt;
&lt;li&gt;Verify that the task was created.&lt;/li&gt;
&lt;li&gt;Mark the task as completed.&lt;/li&gt;
&lt;li&gt;Confirm that exactly one task was completed.&lt;/li&gt;
&lt;li&gt;Verify that the counter displayed "0 items left."&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;&lt;strong&gt;Result: PASSED.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;PROOF recorded the execution, verified the expected conditions, and generated a report containing the evidence.&lt;/p&gt;

&lt;p&gt;I also built a separate demonstration involving a broken contact form. PROOF detected a validation error, and a second execution verified the corrected behavior.&lt;/p&gt;

&lt;h2&gt;
  
  
  🛠️ The technology
&lt;/h2&gt;

&lt;p&gt;The prototype uses:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;React and TypeScript for the interface.&lt;/li&gt;
&lt;li&gt;Node.js for the local verification service.&lt;/li&gt;
&lt;li&gt;Playwright with Chromium for browser automation.&lt;/li&gt;
&lt;li&gt;HTML reports with embedded screenshots.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal is to make test results easier to inspect, understand, and share.&lt;/p&gt;

&lt;h2&gt;
  
  
  ⚠️ What PROOF cannot do yet
&lt;/h2&gt;

&lt;p&gt;I want to be transparent about its current limitations.&lt;/p&gt;

&lt;p&gt;PROOF is still an early prototype that runs locally.&lt;/p&gt;

&lt;p&gt;It doesn't automatically discover every bug.&lt;/p&gt;

&lt;p&gt;Tests require configured interactions and expected outcomes.&lt;/p&gt;

&lt;p&gt;It currently supports authorized local applications rather than arbitrary public websites.&lt;/p&gt;

&lt;p&gt;And a passing test proves only the specific conditions that were checked.&lt;/p&gt;

&lt;h2&gt;
  
  
  🚀 Where I want to take it
&lt;/h2&gt;

&lt;p&gt;I'm exploring whether PROOF could become a useful product for:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Developers building applications with AI.&lt;/li&gt;
&lt;li&gt;Freelancers delivering websites to clients.&lt;/li&gt;
&lt;li&gt;Small development agencies.&lt;/li&gt;
&lt;li&gt;Teams that need visual evidence before releasing software.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;My goal is to make software verification more accessible, especially for people who build applications quickly with AI-assisted tools.&lt;/p&gt;

&lt;h2&gt;
  
  
  💬 I'd love your feedback
&lt;/h2&gt;

&lt;p&gt;I'm currently validating the idea before investing in a larger product.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;If you build applications with AI, how do you verify that they actually work before shipping?&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Would you use a tool that automatically runs your configured tests and creates shareable, evidence-backed reports?&lt;/p&gt;

&lt;p&gt;And what would it need to offer for you to consider paying for it?&lt;/p&gt;

&lt;p&gt;I'm also looking for a few developers interested in trying an early version.&lt;/p&gt;

&lt;p&gt;Let me know what you think!&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Disclosure: PROOF was developed with AI-assisted coding tools. This article was also prepared with AI writing assistance and reviewed against the prototype's demonstrated capabilities.&lt;/em&gt;&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fh1k15wtmacjlnrft4koy.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fh1k15wtmacjlnrft4koy.png" alt=" " width="800" height="622"&gt;&lt;/a&gt;&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fqgg4rhgen9tsrfv6z5yc.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fqgg4rhgen9tsrfv6z5yc.png" alt=" " width="800" height="622"&gt;&lt;/a&gt;&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>webdev</category>
      <category>testing</category>
      <category>ai</category>
    </item>
  </channel>
</rss>
