I Needed an APP to Track My Learning Journey, AI built it in less than half an hour with a single prompt

Source: Dev.to

Overview

I spent several months trying to build a Learning Hub App using various tools, only to end up in a cycle of fixing one thing and breaking another. This time I turned to the Goose coding agent and Claude Opus 4.5 with a single, clear prompt—no .prd file required. The result was a fully functional app built, tested, and ready to use in under half an hour.

Prompt Used

I would like you to build me an app so that I can easily manage URLs for blog posts, podcasts, videos and other things that I would like to learn when it comes to AI. It would be great to be able to easily add the URL and then have a title and description field which can be populated when adding it. Search by category would be great. I would be cool to have some sort of system like a todo list so when it is done it goes to a different place but is still findable should I want to share it with someone. Maybe even notes so I could add some notes on it for later findings or note taking. Should be able to prioritize things so that I learn things based on a particular order—maybe drag‑and‑drop so I can change it. It should be a fun app that I can easily deploy, nice and easy on the eye. It would also be great to have a section where I can put ideas on content creation based off of the stuff I have learned (e.g., blog posts, videos, etc.). Can you come up with a plan for this?

Generated Plan

The agent responded with a comprehensive plan titled “AI Learning Hub”, covering:

• Core Features Overview
• Tech Stack Recommendation
• Data Structure
• UI Layout Concept
• Build Phases
• Future Improvements (nice‑to‑have features)

After a couple of minor adjustments, the agent proceeded to implement the plan.

Build Process

• Implementation: The app was generated automatically by the agent.
• Testing: Using Playwright MCP, the agent ran end‑to‑end tests while I watched. It added links, exercised filters, and interacted with the UI just like a real user, confirming that everything worked.

Features Implemented

• URL management for blog posts, podcasts, videos, etc.
• Automatic title and description extraction when a URL is added.
• Category‑based search.
• Todo‑style workflow: items move to a “completed” section but remain searchable.
• Notes field for each entry.
• Prioritization with drag‑and‑drop reordering.
• Simple, eye‑friendly UI that can be deployed quickly.
• Idea board for future content creation (blog posts, videos, etc.).

Adding URLs

To add new resources, simply paste the URLs into the Goose interface. The configured Playwright MCP will fetch the metadata and populate the app automatically—no manual coding required.

Reflections

• I never opened an editor or inspected the generated code. The app works as intended, which is all I needed for this project.
• While I care about code quality, I’m comfortable trusting the LLM and agent to produce clean, standards‑compliant code for now.
• I plan to add additional tests and performance checks later, and I may review the code during a pull‑request, but the speed and ease of this workflow have truly impressed me.

Try It Yourself

If you want to replicate this experience, start with the prompt above and let a capable coding agent (e.g., Goose + Claude Opus 4.5) handle the rest. Enjoy building your own AI‑focused learning hub in minutes!