Skip to content

Build your own personal finance analytics using Plaid, Google Sheets and CircleCI.

License

Notifications You must be signed in to change notification settings

automatedmachine/build-your-own-mint

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Build Your Own Mint

Important Disclaimer

All this repo does is talk to Plaid/Google APIs and write tokens to your local file system. If you don't feel safe entering real bank credentials, audit the code yourself to make sure.

Install Dependencies

This project uses Node.js. Run npm install in the repo root to install necessary dependencies.

Setting up API keys

First things first - rename .env.sample to .env. Variables in this file will be loaded as environment variables. This file is ignored by Git.

Plaid

  • You will first need to sign up for Plaid and apply for the development plan. You might need to wait for a day or two to get approved. It's free and limited to 100 items (i.e. banks), so it should be more than enough for your personal use.

  • Once approved, fill out the following in .env:

    • PLAID_CLIENT_ID
    • PLAID_SECRET
    • PLAID_PUBLIC_KEY
  • Now you need to connect to your financial institutions to generate access tokens.

    Run npm run token-plaid <account> where account is an id for the bank you want to connect (it's for your personal reference, so you can name it anything). This will start a local server which you can visit in your browser and go through the authentication flow. Once you've linked the bank, its associated access token will be saved in .env.

    This process needs to be repeated for each bank you want to connect. Make sure to run each with a different account name.

  • If you've done everything correctly, running npm run test-plaid now should log the recent transactions in your connected accounts.

Google Sheets

I use a Google Sheet because it's convenient. If you don't trust Google or want to build your own fancy interface, you can totally do that - but that's out of scope for this demo.

  • First, create a Google Sheets spreadsheet, and save its ID in .env as SHEETS_SHEET_ID.

  • Then, go to Google Sheets API Quickstart, and click the "Enable the Google Sheets API" button. Follow instructions and download the credentials JSON file. Take a look at the file and fill in the following fields in .env:

    • SHEETS_CLIENT_ID
    • SHEETS_CLIENT_SECRET
    • SHEETS_REDIRECT_URI (use the first item in redirect_uri)
  • Run npm run token-sheets. This will prompt for auth and save the token in .env.

  • Now run npm run test-sheets. You should see your sheet's cell A1 with "It worked!".

Transform your Data

  • With the APIs sorted out, now it's time to connect them. Open lib/transform.js - this is where you can write your own logic to map incoming transactions to cell updates. How to structure the spreadsheet to use that data is up to you.

  • By default, the transaction date range is from the beginning of last month to now. You can adjust this in lib/fetch.js.

  • Once you've setup your own transform logic, run node index.js. If everything works, your spreadsheet should have been updated.

  • This repo only handles transactions, but it should be pretty straightforward to add balances. (logic for fetching balances is in fetch.js already)

Automate the Updates

The repo contains a CircleCI config file which runs the update every day at 5AM UTC (midnight US Eastern time). You can adjust the cron config to tweak the time/frequency of the updates. Note that your local .env is not checked into the repo, so you will need to copy all those env variables into your CircleCI project settings.

This is totally optional if you don't trust CI with your tokens. Just run it manually when you want to update things.

About

Build your own personal finance analytics using Plaid, Google Sheets and CircleCI.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 48.5%
  • JavaScript 43.5%
  • CSS 8.0%