Skip to content

Latest commit

 

History

History
236 lines (178 loc) · 5.71 KB

File metadata and controls

236 lines (178 loc) · 5.71 KB

🚀 READY TO USE

Your personal finance system is production-ready. Here's how to get started with your real data.

📤 Step 1: Share Your Bank CSV Formats

I've built parsers for:

  • ✅ Chase
  • ✅ American Express
  • ✅ Bank of America
  • ✅ Discover
  • ✅ Capital One
  • ✅ Wells Fargo
  • ✅ Citi
  • ✅ Generic (fallback for any bank)

If Your Bank Isn't Listed

Share 3 rows from each bank's CSV export:

Date,Description,Amount,Balance
01/15/2024,GROCERY STORE,-45.67,1234.56
01/16/2024,PAYCHECK,2000.00,3234.56

I'll build a custom parser in ~5 minutes. See CSV-FORMAT-GUIDE.md for details.

🎯 Step 2: Setup & Import

# 1. Install dependencies
pnpm install

# 2. Setup database
cd packages/database
pnpm prisma generate
pnpm prisma migrate dev --name init
tsx prisma/seed.ts
cd ../..

# 3. Create .env
cp .env.example .env

# 4. Create accounts (one per bank/card)
pnpm cli accounts add
# Repeat for each account

# 5. Import transactions
pnpm cli import chase-checking.csv
pnpm cli import amex-card.csv
# System auto-detects format

# 6. Review and categorize (trains the system)
pnpm cli review

# 7. Detect subscriptions
pnpm cli subscriptions detect
pnpm cli subscriptions hidden

# 8. Generate financial snapshot
pnpm cli snapshot -o my-finances.md

💬 Step 3: Talk to Cursor

Open my-finances.md in Cursor and ask:

🆘 Emergency Mode

"I'm in deep financial shit. Net worth is $X, debt is $Y, spending $Z more than I make. 
Create a complete 6-month recovery plan. What do I cut first? How do I prioritize debt? 
Be specific with dollar amounts and timeline."

💰 Subscription Audit

"Show all my subscriptions sorted by cost. Which ones should I cancel based on value? 
Calculate yearly savings if I cut the bottom 50%. Be brutal about what's a waste."

📊 Spending Analysis

"Where am I hemorrhaging money? Show top 5 problem categories. What are the easiest 
cuts that save the most? Give me a realistic budget based on my actual income."

🔍 Hidden Costs

"Find all the 'death by a thousand cuts' - small recurring charges under $20/month. 
Show total yearly cost. Recommend what to cancel."

📈 Portfolio Analysis

"Analyze my investment allocation. Am I diversified properly? Should I rebalance? 
What specific trades should I make?"

💡 Income vs Expenses

"Based on my spending, what's my REAL required monthly income? Can I negotiate any 
fixed costs down? Should I focus on earning more or spending less?"

🎯 Goal Planning

"I want to save $X in Y months for [goal]. Based on my cash flow, is this realistic? 
What specific budget changes do I need?"

😤 Tough Love

"Stop being nice. Look at my finances and tell me hard truths. What am I doing that's 
financially stupid? What lifestyle changes do I NEED to make?"

🔄 Monthly Workflow

# Export new transactions from banks
# (download CSVs)

# Import
pnpm cli import new-chase-Jan.csv
pnpm cli import new-amex-Jan.csv

# Quick review
pnpm cli review

# Update subscriptions
pnpm cli subscriptions detect

# Generate monthly report
pnpm cli snapshot -o monthly-jan.md

# Compare to last month in Cursor
"Here's January vs December. Am I improving? What changed? 
Am I on track with the recovery plan?"

🛠️ Advanced Usage

Check Supported Formats

pnpm cli formats

API Server (optional)

pnpm api
# Available at http://localhost:3000

Database Studio

pnpm db:studio
# Visual database explorer

Find Hidden Subscriptions

pnpm cli subscriptions hidden --max 25
# Find recurring charges under $25

Export for Tax/Accounting

pnpm cli snapshot -o taxes-2024.md -f json
# JSON format for processing

📊 What You Get

The financial snapshot includes:

Net Worth - Total assets minus liabilities
Cash Flow - Income vs expenses (30/90/365 days)
All Subscriptions - With monthly/yearly totals
Spending by Category - Top expense categories
Investment Portfolio - Holdings and performance
Trends - Month-over-month changes
Data Quality Score - How complete your data is

🎓 Pro Tips

  1. Import regularly - Weekly or bi-weekly keeps data fresh
  2. Review immediately - Categorize while transactions are fresh in memory
  3. Train the system - Manual corrections make auto-categorization smarter
  4. Monthly snapshots - Track progress over time
  5. Be honest with Cursor - Real problems get real solutions

🔒 Privacy Notes

  • ✅ All data stays local (SQLite database)
  • You generate snapshots for AI analysis
  • No telemetry or external API calls
  • You control what data leaves your machine

📤 Next Steps

Right now:

  1. Share your bank CSV formats (if not already supported)
  2. I'll build custom parsers
  3. You import real data
  4. Start getting AI-powered insights

When ready:

  • Compare to Copilot (import same data, compare accuracy)
  • Track month-over-month improvement
  • Use for tax planning
  • Financial decision making

🆘 Need Help?

Format issues? → Share CSV sample
Import errors? → Check CSV-FORMAT-GUIDE.md
Categorization wrong? → Run pf review to train system
Data quality low? → Fix uncategorized transactions


You now have:

  • ✅ Intelligent multi-bank CSV importer
  • ✅ Learning categorization system
  • ✅ Subscription detection (including hidden costs)
  • ✅ Comprehensive financial snapshots
  • ✅ LLM-ready reports for AI analysis
  • ✅ Production-grade data quality

Ready to import your bank CSVs and get started?

Share your CSV formats and let's make this work with YOUR real financial data.