Debug my life

Inspiration Every developer knows the feeling: you stare at a bug, unsure what's wrong, and then you start running through a mental checklist. Print statements, breakpoints, bisecting the problem. It works. So I asked myself: what if we applied that same structured debugging process to life decisions? We all get stuck on big choices. Career moves, relationships, finances. And most of us just spin in circles, weigh the same two points over and over, then either overthink into paralysis or make a gut call and hope for the best. There's no framework, no process, no log of what happened afterward. Debug My Life treats life decisions like software bugs. You file a bug report, run diagnostics, commit a decision, and later write a retrospective. It turns messy, emotional decision-making into something structured without stripping out the human element. What it does Debug My Life is a full-stack personal decision journal built around four core actions: File a Bug Report - Describe the decision you're stuck on, set a priority level (P0/P1/P2), tag a category (Career, Finance, Health, etc.), and add context. The conversational multi-step form makes it feel like talking through the problem, not filling out a spreadsheet. Run Diagnostics - Work through four structured thinking frameworks:

Pros & Cons Matrix with weighted scoring so not all factors count equally Regret Minimization where you imagine being 80 and looking back Second-Order Effects that map out consequences 2-3 levels deep Pre-Mortem where you imagine each option failing and ask why

Commit Your Decision - Lock in your choice with a rationale and confidence score (1-10). Once committed, diagnostics are frozen so you can't retroactively edit your reasoning. Write a Retrospective - After 24+ hours, revisit the decision: what actually happened, what you'd do differently, and rate past-you's call on a 5-star scale. The Git Log gives you a timeline of every decision you've ever made, filterable by category and status. The Diff View analyzes your patterns over time, surfacing insights like "Your career decisions average 4.2/5 in retrospectives" or "You've filed 3 finance bugs but only committed 1." How I structured conversations with MeDo I started with a detailed requirements document covering all 7 pages, the database schema (8 tables), business rules, and the full design system. Instead of prompting MeDo with vague descriptions, I gave it structured specs for each page including the exact conversational flow for the bug report form, the four diagnostic frameworks with their specific prompts, and the status state machine (open > diagnostics > committed). The key insight was being explicit about what not to build. I listed features excluded from v1 (dark/light toggle, export, AI recommendations) upfront, which kept MeDo focused on the core loop instead of adding scope. For the diagnostics page specifically, I described each framework as a self-contained section with its own input types (sliders, text areas, two-column layouts, tree structures) rather than one generic "framework" component. This gave MeDo enough detail to generate meaningfully different UIs for each one. The most impressive feature MeDo helped create The diagnostics page. It has four completely different interactive frameworks, each with its own data model and UI pattern, all auto-saving to the database, with a progress tracker that updates in real time. The Second-Order Effects framework in particular, with its branching tree structure that prompts "and then what happens after that?" to go multiple levels deep, came together surprisingly well from a single descriptive prompt. The commit flow is also worth calling out. The "git commit -m 'decision made'" button, the confidence slider, and the way diagnostics lock after commit all work together to create a moment that feels decisive and final, which is the whole point. How I used plugins or APIs to extend functionality The app uses MeDo's built-in database and authentication system for the full-stack backend. User auth with email/password, row-level data isolation (each user only sees their own decisions), and relational data across 8 tables (decisions, pros_cons, regret_minimization, second_order_effects, pre_mortem, commits, retrospectives, users) are all handled through MeDo's generated backend. The Diff View analytics page computes pattern insights dynamically from the user's data using template-based analysis queries rather than hardcoded text.

Built With

• css
• html
• javascript
• medo
• supabase