Capstone: ship Smart Ledger
Assemble the product
This capstone starts empty—there is no fake ledger. A real user adds their own transactions. The app validates and normalizes input, stores amounts as agorot, persists locally, derives KPIs, filters rows, deletes entries, and exports a JSON backup.
The export is a real artifact. The storage is real browser storage. The only limitation is deliberate: this learning project never connects to a bank and never asks for banking credentials.
Instructions
Implement loadTransactions and saveTransactions using localStorage and JSON.
Hint: Restore [] on parse failure; stringify transactions on every mutation.
Where does this go?
// Assemble: load/save, validation, add/delete, derived KPIs, filter, and JSON export.Write it on line 5 — under that comment, after the code you already added.
Validate and normalize input in createTransaction.
Render KPIs and filtered rows from the transactions array.
Implement add and delete flows, saving after each mutation.
Export a real JSON backup with Blob and URL.createObjectURL.
Your Code
Describe what you want to build for this step and I'll write the code — then click Apply to drop it into your editor. This step: Implement loadTransactions and saveTransactions using localStorage and JSON.