dinero

Tools and scripts to manage my personal finances. Your own Mint/YNAB, self hosted.

• Plaid for collecting transactions from financial institutions
• SQL Database for storing transactions
  • I personally use Postgres but any SQLAlchemy compatible DB should work
  • This README uses SQLite
  • I use NocoDB as a viewer/explorer with filters and groups by Account, Category and so on
  • I use Metabase to create dashboards
  • Your imagination is the limit here!
  • You can read more about how I use this + some screenshots in my wiki.
• Inspired by yyx990803/build-your-own-mint.

Requirements

Plaid

• Create a Plaid account
• You need development access to handle multiple accounts
  • You can use it on development mode for free up to 100 accounts
• I asked for production access and it was simple, just fill the form and say you are not a company and that you will use it for personal use
• Get your client_id and secret from the Plaid dashboard

Logging in to Banks

Use the plaid/quickstart to login to your institution.

Once you've linked the bank save the ACCESS_TOKEN. in the config file. See settings below.

Database

Simply write the SQLAlchemy connection string in the config file.

Config

The project reads it's config from ~/.config/dinero/config.toml. You can see a sample in config.sample.toml.

For Plaid you need to set the token for each institution:

[plaid.tokens]
bank_1 = "access-development-XXXXXXXXXXXXXXXX"
bank_2 = "access-development-YYYYYYYYYYYYYYYY"

And a mapping to make the Account ID human readable:

[plaid.account_id_to_name]
THIS_IS_A_LONG_ID_1_XXXXXXXXXXXXXXXXXXXXXX = "Bank 1 Checking"
THIS_IS_A_LONG_ID_2_YYYYYYYYYYYYYYYYYYYYYY = "Bank 2 Credit Card"

Usage

After you have the requirements and config file.

Install by cloning the repo and installing the dependencies using your favorite Python environment manager.

Create DB and tables:

dinero init-db

Get new transactions and add them to the database:

dinero transactions

Example output:

2023-09-21 11:06.45 [info     ] Transactions downloaded        name=XXX records=49
2023-09-21 11:06.45 [info     ] Transactions pending           name=XXX records=1
2023-09-21 11:06.45 [info     ] Transactions not pending       name=XXX records=48
2023-09-21 11:06.58 [info     ] Loaded table                   recods=662 year=2023
--------------------------------------------------------------------------------
My Bank - Checking
--------------------------------------------------------------------------------
2023-09-21 11:06.58 [info     ] Queried transactions           n=17
2023-09-21 11:06.58 [info     ] Pending transactions           n=0
2023-09-21 11:06.58 [info     ] New transactions to be added   n=2
2023-09-21 11:06.58 [info     ] Transaction                    transaction={'Account': 'BoA Checking', 'Amount': XXX, 'Category': '', 'Date': '2023-09-20', 'Description': 'XXXX', 'Subcategory': ''}
2023-09-21 11:06.58 [info     ] Transaction                    transaction={'Account': 'BoA Checking', 'Amount': XXX, 'Category': '', 'Date': '2023-09-20', 'Description': 'XXXX', 'Subcategory': ''}
2023-09-21 11:06.58 [info     ] Existing transactions          n=15
2023-09-21 11:06.58 [info     ] Error transactions             n=0

================================================================================
All accounts summary from 2023-08-22 to 2023-09-21:
================================================================================
Using Table for year: 2023
Records in table: 662
Transactions queried: 49
Pending transactions: 1
Transactions analysed (pending False): 48
New records to be inserted: 5
Existing transactions: 43
Errors transactions (not valid dates for table): 0

Insert transactions to the Table? [Y/n]

Generate a set of simple rules that will be used to categorize transactions:

just rules

Generate a dataset with all transactions in CSV and SQLite:

just dataset

Other

There is a handy function to generate a Pandas DataFrame with all transactions: I use this to do some analysis in a Jupyter Notebook.

from dinero import analysis
df = analysis.get_dataframe()

Contributions

While I am happy to accept any contributions this is 100% tailored to how I use it so I might reject stuff I won't use.

If you want to do other things such as connecting to Airtable (initially I used Airtable but the new pricing made it not worth it) you should fork this repo and make your own changes.