Prompt para analizar tus finanzas personales con Claude Code. Parsea extractos bancarios (Santander u otros), categoriza transacciones, genera un dashboard interactivo con Chart.js y te da recomendaciones concretas para ahorrar.

financebot_prompt.txt

I have a bank export spreadsheet with my expenses and income from the last year(s). I need you to:

1. **Read and parse** the spreadsheet - it's a bank export (XLS/XLSX) with dates, concepts/descriptions, and amounts. Figure out the format (headers may not be on the first row).

2. **Categorize every transaction** into categories like: Groceries, Dining, Coffee, Nightlife, Events, Bars, Shopping, Subscriptions, Travel, Transport, Car, Health, Beauty, Pets, Home, Fitness, Electronics, Entertainment, etc. Use regex pattern matching on merchant names. Ask me about any merchants you're not sure about
(especially recurring ones).

3. **If I have a separate transfers file**, merge it in - categorize salary, investments, rent, savings, etc.

4. **Identify fixed/recurring expenses** separately: mortgage, car payments, insurance, community fees, utilities (electricity, water), phone, internet. These should have their own section in the dashboard.

5. **Build an interactive HTML dashboard** (single file, Chart.js from CDN, dark theme) with:
   - Summary cards (monthly salary, **total** spending, savings rate, investments)
     - The savings rate and spending must consider ALL expenses (not just card)
   - Monthly overview chart with: salary, card expenses, fixed expenses, transfers, and net savings
   - Category breakdown (donut chart)
   - Top merchants (horizontal bar chart)
   - Monthly category trends (stacked area)
   - **Fixed spending section**: horizontal bar chart with monthly averages per
     fixed category + stacked monthly evolution chart
   - 50/30/20 rule analysis (Needs vs Wants vs Savings) with donut chart, details panel, and monthly stacked bar
   - Transfers by category (donut chart)
   - Savings rate per month
   - Spending insights: nightlife total, coffee habits, Zalando net (after returns), real savings rate, investment rate, **tax refunds**,
     **investment returns**, grocery monthly, subscriptions, travel, crypto
   - Recommendations section with concrete, data-driven suggestions to reduce spending, with estimated savings per recommendation. Include:
     - Shopping discipline (biggest variable expense)
     - Blind spot analysis (ATM cash + uncategorized transfers)
     - Dining frequency
     - Nightlife budget
     - Subscription audit (check for subscription creep over time)
     - Coffee habits
     - Crypto strategy
     - Summary card with real savings rate and realistic improvement target
   - Searchable/filterable transaction table with pagination

6. **Give me actionable analysis**: how much I spend, how much I save, which categories have room for improvement, seasonal patterns, and specific recommendations with EUR amounts.

7. **Find correlations and patterns**:
   - Day of week spending patterns
   - Payday effect (spending spikes after salary)
   - Subscription creep (are subscriptions growing over time?)
   - Category trends (spending going up or down over time?)
   - Coffee vs nightlife correlation
   - Dining vs groceries substitution effect
   - Shopping spree patterns (clustering of high-spend months)
   - Travel impact on monthly totals
   - ATM withdrawal patterns
   - Month-over-month self-correction (do you compensate after big months?)

Set up a Python venv with requirements.txt (pandas, xlrd, openpyxl) so it's reproducible. Output a categorize.py script that generates a dashboard_data.json, and a dashboard.html that reads it.

Start by reading my file and showing me the structure, then ask me about the biggest merchants you can't identify.