<?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: Warren Wong</title>
    <description>The latest articles on DEV Community by Warren Wong (@somethingwentwell).</description>
    <link>https://dev.to/somethingwentwell</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%2F4236473%2F1ac1cdc1-63a2-4a31-92ee-6dde47d7d8f8.jpg</url>
      <title>DEV Community: Warren Wong</title>
      <link>https://dev.to/somethingwentwell</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/somethingwentwell"/>
    <language>en</language>
    <item>
      <title>Build a human-approved workflow app from a prompt (Jevable request-review walkthrough)</title>
      <dc:creator>Warren Wong</dc:creator>
      <pubDate>Tue, 10 Nov 2026 08:34:22 +0000</pubDate>
      <link>https://dev.to/somethingwentwell/build-a-human-approved-workflow-app-from-a-prompt-jevable-request-review-walkthrough-1d64</link>
      <guid>https://dev.to/somethingwentwell/build-a-human-approved-workflow-app-from-a-prompt-jevable-request-review-walkthrough-1d64</guid>
      <description>&lt;p&gt;I'm the maker of Jevable. This is a technical walkthrough of one shipped example, not a production-readiness or user-count claim.&lt;/p&gt;

&lt;p&gt;Coding agents invent steps easily. Recurring work often needs an accepted graph, a retained run ID, and a human decision before anything continues.&lt;/p&gt;

&lt;p&gt;Jevable is an open-source (MIT) agent skill on GitHub (&lt;code&gt;0.1.0-alpha.1&lt;/code&gt;). An agent can turn a described process into a workflow app you can inspect, test, and run independently. This post covers only the &lt;strong&gt;model-free&lt;/strong&gt; &lt;code&gt;request-review&lt;/code&gt; example. No hosted model. No outbound send.&lt;/p&gt;

&lt;p&gt;Install:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;git clone https://github.com/somethingwentwell/jevable-skill
&lt;span class="nb"&gt;cd&lt;/span&gt; jevable-skill
npm ci --omit=dev
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Needs git, Node.js 22.19+, and npm. The checkout ships prebuilt &lt;code&gt;dist/&lt;/code&gt;, so normal use has no build step. Point your agent at that checkout as the skill location.&lt;/p&gt;

&lt;p&gt;The example routes on an input boolean called &lt;code&gt;needsReview&lt;/code&gt;:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;needsReview: false&lt;/code&gt; → the routine path completes (exit &lt;strong&gt;0&lt;/strong&gt;)&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;needsReview: true&lt;/code&gt; → Human Approval waits (exit &lt;strong&gt;3&lt;/strong&gt;); do &lt;strong&gt;not&lt;/strong&gt; approve just to “finish” the demo&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;code&gt;needsReview&lt;/code&gt; is authored in the workflow. It is not automatic sensitivity detection.&lt;/p&gt;

&lt;p&gt;You should see one completed run, one waiting approval with retained run IDs, and no automatic approve/reject.&lt;/p&gt;

&lt;p&gt;Limits: early alpha — not production-ready; hosted model providers remain unverified in this public package; custom code nodes validate but do not execute; some public test helpers / boundary tooling are reconstructions, not private-baseline byte matches.&lt;/p&gt;

&lt;p&gt;If you try the clone path, comment with your OS/Node version and the first error you hit.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Repo:&lt;/strong&gt; &lt;a href="https://github.com/somethingwentwell/jevable-skill" rel="noopener noreferrer"&gt;https://github.com/somethingwentwell/jevable-skill&lt;/a&gt;&lt;br&gt;&lt;br&gt;
&lt;strong&gt;License:&lt;/strong&gt; MIT&lt;br&gt;&lt;br&gt;
&lt;strong&gt;Earlier Cursor Forum thread:&lt;/strong&gt; &lt;a href="https://forum.cursor.com/t/jevable-an-mit-agent-skill-for-runnable-workflows-with-human-approval/177901" rel="noopener noreferrer"&gt;https://forum.cursor.com/t/jevable-an-mit-agent-skill-for-runnable-workflows-with-human-approval/177901&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Disclosure: I used AI assistance to draft and tighten this article; I reviewed the technical claims and am responsible for the content.&lt;/p&gt;

</description>
      <category>cursor</category>
      <category>agents</category>
      <category>workflow</category>
      <category>node</category>
    </item>
  </channel>
</rss>
