weekly-figureslisted
Install: claude install-skill sandeep-sidhu/claude-business-skills
# Weekly figures
Owners of small businesses usually know how busy they were, not how they did. A Monday page that answers "sales, cash, who owes us" in two minutes changes decisions: who to chase, whether to hire, whether that quiet week was a blip. It only works if the numbers are trusted, so every figure here is calculated by `scripts/figures.py` and points back to the rows it came from. Your job is the reading: what changed, and what to look at.
## What you need
Any one or more of these as CSV (the script matches column names loosely):
- **Sales**: an invoice list or sales day book, with a date and a net or total column. A type column lets credit notes count as negative.
- **Debtors**: an aged debtors or unpaid invoices report, with customer, outstanding amount, and a due date or invoice date.
- **Bank**: a bank statement or bank transactions export, with a date and either a signed amount or money-in / money-out columns.
At least five weeks of sales and bank history makes the 4-week average real; less still works and the script says how many weeks it used. `references/exports.md` lists which report to export from each package.
## Steps
1. **Run the script** on whatever they gave you:
```bash
python3 scripts/figures.py --sales sales.csv --debtors debtors.csv --bank bank.csv --today YYYY-MM-DD --format json
```
Weeks run Monday to Sunday; by default the report covers the last full week before `--today`. Use `--week-ending` (a Sunday) for a different week, `--term