The Revisiting Tool

Set it, forget it, and rediscover it. The Revisiting Tool is the simplest way to get automatic email reminders for any link. Your personal digital memory assistant.

Comment

Inspiration

The idea for The Revisiting Tool was born from a combination of personal frustration and a spark of public validation. Like many people, my browser is a digital graveyard of good intentions—saved articles, must-watch videos, and crucial reports that I promise myself I'll "get back to," but never do. My bookmark bar is a black hole, and "read-it-later" services quickly become "read-it-never" archives.

The real "aha!" moment came while I was scrolling through X. I saw a post where someone expressed a simple wish: they wanted a button on X that would let them "revisit this post in x days." It was a brilliantly simple concept that perfectly captured the fleeting nature of online content. That single post crystallized my own frustration into a clear product idea: what if this feature existed for the entire internet?

When I saw the Bolt.new "One-Shot" challenge, everything clicked. The challenge wasn't just a constraint; it was the perfect opportunity. Could I take this simple, powerful idea—validated by a real user's wish—and build a complete, full-stack solution with just a single, meticulously crafted prompt? The challenge was on.

What it does

The Revisiting Tool is a focused web application that ensures you never lose track of important online content. Its core functionality is simple and powerful:

  1. Capture: A user saves any URL—an article, a video, a social media post—along with a title and an optional description.

  2. Schedule: They attach a future date to the saved link, scheduling when they want to be reminded.

  3. Remind: On the scheduled date, the app's automated system sends a clean, actionable email directly to the user's inbox containing the link and its details.

This simple loop turns a passive "read-it-later" list into an active, reliable pipeline for discovery and learning, bringing valuable content back to you exactly when you need it.

How I built it

This project was built entirely for the Bolt.new "One-Shot Competition," meaning the whole application was generated from a single, detailed prompt. My role was that of a system architect, designing a complete blueprint for Bolt.new to execute.

  • Tech Stack: React, Tailwind CSS, Supabase, and the Resend API.

  • The Prompt: I crafted a comprehensive prompt that specified:

    • The complete database schema for profiles and reminders tables in Supabase.
    • Full user authentication flows using Supabase Auth.
    • The complete backend automation logic, including a Supabase Cron Job that triggers an Edge Function.
    • The Edge Function to send emails via Resend, including logic to handle their API rate limits.

  • Final Steps: After Bolt.new generated the code, I added my Resend API key, added the sender email address in the Edge Function, and scheduled the daily cron job in the Supabase dashboard.

Challenges I ran into

While the AI generation was highly successful, bringing the project to life required debugging three specific issues to make the generated code fully functional:

  1. Edge Function Query: The initial database query in the email notification function was malformed. I corrected the query syntax to establish the proper relationship.

  2. Authentication State: The app had a session handling bug that caused an infinite loading state when a user refocused the browser tab. This was resolved by refining the client-side state management logic.

  3. Deployment Routing: On Netlify, the app faced a standard single-page application (SPA) routing issue. This was fixed by adding a _redirects file to correctly handle client-side routes on page refresh.

Crucially, these were all error corrections as permitted by the rules. No new functionality, features, or UI were added or edited, ensuring the project remained a true "One-Shot" submission.

Accomplishments I am proud of

  • Successfully Completing the One-Shot Challenge: Building a full-stack, production-ready application with user auth, a database, and automated background jobs from a single prompt is my biggest accomplishment here. It proves the power of this new development paradigm.

  • Creating a Genuinely Useful Tool: This isn't just a tech demo; it's a solution to a problem I and many others face daily. It's immensely satisfying to have built a tool I will personally use.

  • Implementing a Complex Backend Flow: The automated email notification system—which involves a scheduler, a serverless function, and a third-party API—is a complex piece of engineering. Having this entire workflow generated correctly and function perfectly is something I'm incredibly proud of.

What I learned

  1. Prompting is the New Architecture: The developer's role is evolving. For this project, my most valuable skill wasn't writing lines of code, but architecting a system and translating that architecture into a comprehensive, detailed prompt.

  2. Serverless is Powerful: This project reinforced how incredibly powerful and cost-effective a serverless stack can be. The ability to run automated jobs and send emails without managing a single server is a game-changer for solo developers and small teams.

  3. The "Last Mile" Still Matters: AI can build the car, but a human still needs to put the key in the ignition. Understanding how to configure the environment, set up schedulers, and debug the integration between services remains a critical developer skill.

What's next for The Revisiting Tool

The Revisiting Tool is currently a fully functional web app, but there's a clear path forward to make it even more indispensable.

  • Browser Extension: A one-click browser extension to save links without ever leaving the current page is the most logical next step.

  • More Notification Channels: Adding options for push notifications, or even integrations with Slack and Discord, would give users more control over how they are reminded.

  • Organizational Features: As users add more reminders, they will need tools to organize them, such as tags, folders, or categories.

  • Smarter Reminders: I envision a future version that uses AI to suggest optimal times to revisit content based on its type (e.g., suggesting spaced repetition for learning materials).

  • AI Follow-Up Agent: A powerful future feature where an AI agent automatically researches the topic on the reminder's due date. For example, if you save a post where someone predicts a future event, the reminder email could include an AI-generated summary of whether that prediction actually came true.

Built With

Share this project:

Updates