Back to all 15 projects
Project 15 / 15Deployed
Private source

MoatMine

Risk-aware investing: every stock gets a graded safety score, every buy is checked against your own risk rules, and the portfolio is stress-tested against past crises.

Core stack
  • Next.js
  • TypeScript
  • Python
  • FastAPI
  • PostgreSQL
  • Tauri
Source access

Source is private.

Chapter 01

Overview

Most investing apps make buying easy and leave the risk to the user. MoatMine puts the risk first: a short questionnaire sets a risk profile with editable rules — maximum position and sector size, minimum safety score, volatility and loss limits — and every planned buy is checked against them, passing, warning (with a written reason and a cooling-off wait) or blocking. Each stock gets a 0–100 safety score with A–F grades by area, including a moat grade built from up to ten years of return on capital, margin stability and growth. The portfolio is replayed through the dot-com crash, 2008, COVID and 2022 against the user's loss limit. One Next.js UI serves both the web and a Tauri desktop app, backed by one FastAPI service.

Chapter 02

Highlights

  • Safety score with A–F grades for financial health, earnings quality, price risk, crisis resilience, size, valuation and moat
  • Every buy is checked against the user's own rules: pass, warn with a written reason and cooling-off wait, or block
  • Portfolio stress tests replay current holdings through six historical crises and 10/20/35% market falls
  • Eight fair-value models side by side — DCF, Graham, Lynch, dividend discount, EPV and multiples — with every assumption editable
  • Tool-using AI agent routed through ModelDispatcher, bound to the user's risk profile and with no tool that places trades
  • The same tools exposed over MCP with per-user revocable tokens