# Overwhelmed by Overengineering in Project Tracking Tools. The Result? I Built a Lightweight, Local-First Project Tracker.

> Source: <https://dev.to/aiexplorerai/overwhelmed-by-overengineering-in-project-tracking-tools-the-result-i-built-a-lightweight-2pmp>
> Published: 2026-06-15 05:25:05+00:00

Hey everyone 👋

I want to give a massive shoutout to AI — it has injected a whole new level of passion and joy into coding for a non-tech guy like me.

Over the last few weeks, I’ve been exploring what AI can do. I was learning fast, building small experiments, testing ideas… and then I hit a problem:

👉 I completely lost track of what I was building and learning.

The Problem: I tried the usual approaches:

Word documents

Excel sheets

They worked… but felt clunky and disconnected.

Then I looked at modern project management tools — and honestly, they felt massively overengineered for a single developer:

I didn’t need a company-grade system.

I needed something simple, fast, and private.

The Idea - I wondered:

What if I used AI to build the exact tool I wish existed—for myself?

That, to me, is the real power of AI.

So I built a lightweight, minimalist, local-first project tracker that runs entirely on my own machine.

What I Built

This is a local-first desktop-style web app (served locally) focused on clarity and zero friction.

** Key Features**

Dual-View System

Switch seamlessly between:

An editable spreadsheet-style project grid

A visual Kanban board

Dark Mode

An eye-friendly dark theme powered by CSS variables.

****Rich Developer Diaries

Each project has a dedicated workspace to store:

Step-by-step notes

Copyable code blocks

Architecture decisions

Thought process over time

Zero-Friction Auto-Save

Diary blocks auto-save with an 800ms debounce, quietly working in the background as you type.

🧠 AI Progress Summaries (Optional)

Generates monthly or quarterly markdown summaries of progress using the Claude API from Anthropic.

🔒 Total Data Privacy

Everything lives in a single local SQLite file (projects.db).

No cloud sync. No browser localStorage. Nothing that disappears after a cache clear.

🏗️ **Tech Stack (Intentionally Simple)**

I wanted zero setup friction and maximum portability, so I avoided frontend frameworks, npm, and bundlers entirely.

Backend: Python + Flask (lightweight local HTTP server)

Database: SQLite

Write-Ahead Logging (WAL) enabled

Data stored as JSON blobs → no migrations needed when adding fields

Frontend:

Semantic HTML5

Vanilla CSS3

Vanilla JavaScript using native fetch()

No magic. No heavy abstractions. Just simple, understandable code.

**Open Source**

The project is fully open source, and I’m genuinely proud of how it turned out.

GitHub Repository:[https://github.com/AIExplorer-ai/project-tracker?utm_source=chatgpt.com](https://github.com/AIExplorer-ai/project-tracker?utm_source=chatgpt.com)

[Project Tracker on GitHub](https://github.com/AIExplorer-ai/project-tracker?utm_source=chatgpt.com)

Feedback Welcome 🙏

I’d love to hear:

What you think of the approach

Whether you face similar tracking problems

Any features you think would be worth adding next

Thanks for reading — and thanks to AI for making building fun again 🚀