<?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>Forem: Cathy Liu</title>
    <description>The latest articles on Forem by Cathy Liu (@cathyliusolo).</description>
    <link>https://forem.com/cathyliusolo</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%2F3591953%2Ff7f605f5-eebb-429a-a99c-7b431e58a51c.jpg</url>
      <title>Forem: Cathy Liu</title>
      <link>https://forem.com/cathyliusolo</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://forem.com/feed/cathyliusolo"/>
    <language>en</language>
    <item>
      <title>How I Built an AI Idea Library Using No-Code and Vibe-Code</title>
      <dc:creator>Cathy Liu</dc:creator>
      <pubDate>Fri, 31 Oct 2025 23:44:30 +0000</pubDate>
      <link>https://forem.com/cathyliusolo/how-i-built-an-ai-idea-library-using-no-code-and-vibe-code-1a2b</link>
      <guid>https://forem.com/cathyliusolo/how-i-built-an-ai-idea-library-using-no-code-and-vibe-code-1a2b</guid>
      <description>&lt;p&gt;As a software builder, I’m often stuck on one question: &lt;em&gt;What should I build?&lt;/em&gt; The problem isn’t a lack of ideas — it’s the worry no one will need them after I build. I wanted a library of buildable AI product ideas pulled from real user needs, so I created &lt;a href="https://www.solopreneur.global/i/dev" rel="noopener noreferrer"&gt;solopreneur.global&lt;/a&gt; — for myself and anyone facing the same dilemma.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;My Build Playbook: Steps &amp;amp; Tools&lt;/strong&gt;&lt;br&gt;
I split the work into two parts and spent about 2 weeks in total.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;1. n8n workflow to analyze community conversations&lt;/strong&gt;&lt;br&gt;
The workflow ingests Reddit threads about user pain points related to AI products, filters for strong signals and high engagement, then uses AI to draft complete “idea cards”: original post summary, problem statement, solution outline, core features, business model, and go-to-market notes. The workflow runs automatically each day, so the idea library expands continuously.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. A web app vibe-coded on top of a Next.js boilerplate&lt;/strong&gt;&lt;br&gt;
I first stored identified ideas in a spreadsheet, but it didn’t scale as the volume increased. So, I decided to vibe-code a minimal web app to display, categorize, and continuously publish the growing library.&lt;/p&gt;

&lt;p&gt;Unlike a vibe-coding project built from scratch, we started with a Next.js boilerplate as the foundation. On one hand, it already includes common services—user management, payments, blogging, and analytics—so we just plug in the keys and do a quick test pass. On the other, the boilerplate provides a solid software architecture (clear layering, multilingual routing, and robust support for desktop and mobile, etc.), which lets the AI follow that structure when adding new features, preserving consistency and maintainability.&lt;/p&gt;

&lt;p&gt;For this build I used Claude Code. My preferred workflow is: use Plan Mode to have the AI produce an overall design and implementation plan; once I review and approve it, let the AI implement the entire module end-to-end. The process is highly efficient with almost no rework. &lt;/p&gt;

&lt;p&gt;If you’re curious about Claude Code but haven’t tried it yet, I recommend the official video tutorial: &lt;a href="https://www.deeplearning.ai/short-courses/claude-code-a-highly-agentic-coding-assistant/" rel="noopener noreferrer"&gt;Claude Code: A Highly Agentic Coding Assistant.&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Happy to share more in details, just let me know.&lt;/p&gt;

</description>
      <category>nocode</category>
      <category>sideprojects</category>
      <category>ai</category>
      <category>automation</category>
    </item>
  </channel>
</rss>
