An 'almost non-opiniated' open and free MVC architecture to create your own AI bot:
- MODEL your data any way you like. Default set up includes finance, goal, time and run management CRUD modules.
- VIEW is the GNU-LINUX terminal
- CONTROLLER is a Function Calling AI Vendor. Default set up includes OpenAI integration
The bot is designed with an MVC framework, that leverages OpenAI function calling as the controller, to make it more responsive to user inputs. For instance, the rgw commnad (these are my initials, you can name it whatever you want. see installation section), followed by an organic user input can perform CRUD operations on the database:
FINANCE MODULE
rgw "add expense - spent INR 500 on coffee"
rgw "add expense - spent INR 700 on Dining yesterday"
rgw "update expense - update expense id 5 - the cost was INR 450"
rgw "delete expense id 5"
rgw "show me my expenses"
rgw "show me my expenses in the last 46 to 78 days"
rgw "show me my expenses in the last 46 to 78 days cumulatively"
rgw "show me my expenses in the last 46 to 78 days cumulatively and exlcude debt"
rgw "show me my expenses line chart in the last 46 to 78 days"
rgw "show me my expenses line chart in the last 46 to 78 days cumulatively"
rgw "show me my expenses line chart in the last 46 to 78 days cumulatively and exclude debt"
RUNS MODULE
rgw "log run - pre run weight is 205.7 lbs, post run weight is 202.3 lbs, temperature is 88 F, distance is 2.67, fat burn zone is 22 mins, cardio zone is 1 min"
rgw "update run id 14 - distance is 3.6"
rgw "delete run id 14"
rgw "show me my runs"
rgw "show me my running weight line chart"
rgw "show me my running fat burn zone line chart"
rgw "show me my running fat burn zone line chart for the last 14 to 37 days"
rgw "show me my running fat burn zone line chart for the last 14 to 37 days cumulatively"
rgw "show me my running distance line chart"
rgw "show me my running distance line chart for the last 14 to 37 days"
rgw "show me my running distance line chart for the last 14 to 37 days cumulatively"
GOALS
rgw "add goal - run a half-marathon by December"
rgw "add action against goal id 2 - persist with low heart rate training for 3 months"
rgw "show me my goals"
rgw "show me my actions"
rgw "show me my timesheet"
rgw "show me my timesheet of yesterday"
rgw "mark action ids 2, 3, 7 done"
rgw "mark action ids 7, 8, 9 done yesterday"
rgw "show me my actions against goal id 3"
rgw "delete goal id 2"
rgw "delete action id 5"
rgw "update goal id 2 - run a marathon"
rgw "show me my timesheet line chart"
rgw "show me my timesheet line chart for the last 37 to 45 days"
TIME
rgw "show me my events"
rgw "add event - Lunch with Elon Musk on January 1, 2030"
rgw "update event id 34 - Breakfast with Elon Musk on January 1, 2025"
rgw "delete event id 30"
HARDCODED HELPERS
These helpers have been hardcoded (instead of being routed via the OpenAI controller for security purposes, as they interact with senstive data).
List all helpers
rgw --l
Update the bot by pulling the latest version from this git repo
rgw --update
Reset the conversation history of the bot
rgw --reset
Display your passwords, and add, update, delete passwords
rgw --p
rgw --p:create
rgw --p:update
rgw --p:delete
Display your databases + add, update, delete databases credentials, get the schema of your databases, and tunnel into your databases
rgw --db
rgw --db:create
rgw --db:update
rgw --db:delete
rgw --db:schema --id n
rgw --db --id n
Display your virtual machines, add, update, delete virtual machine login credentials, and tunnel into your virtual machines (note: in case you use SSH keys to tunnel in, see step 7 of the installation process below)
rgw --vm
rgw --vm:create
rgw --vm:update
rgw --vm:delete
rgw --vm --id n
STEP I - SET UP THE .ENV
- Rename .env.example to .env.
- In the .env file, add your OpenAI API key, MySQL database (DB) connection details, and credit card coefficient. This coefficient helps estimate a penalty on any purchase made on your credit card to keep track of your total due and outstanding amount.
- You may leave the DEFAULT_DEBT_ID_FOR_PRIMARY_CREDIT_CARD field empty for now, and set it later. This field is required by the default finance module, and helps keep track of how much debt piles up in the main credit card you use.
- If you're using Google Cloud MySQL, add your server's/machine's IP address to Google Cloud SQL's allowed addresses.
STEP II - INSTALL PYTHON AND MYSQL
- Install mysql and python
sudo apt-get update
sudo apt-get install python3
sudo apt-get install mysql-server
- Use MySQL command-line to create your bot's database and tables using the provided commands. The below set up works well with the default functions of in services/modules. Feel free to make it your own.
mysql -h YOUR_DB_HOST_IP -u YOUR_DB_USERNAME -p
CREATE DATABASE `<database_name>`;
USE `<database_name>`;
CREATE TABLE actions (
id INT PRIMARY KEY AUTO_INCREMENT,
goal_id INT,
action VARCHAR(255) NOT NULL,
deadline DATE,
is_active TINYINT(1) DEFAULT b'1'
);
CREATE TABLE cronjob_logs (
id INT PRIMARY KEY AUTO_INCREMENT,
job_description VARCHAR(255),
error_logs JSON,
execution_logs JSON,
executed_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP
);
CREATE TABLE debt_payments (
id INT PRIMARY KEY AUTO_INCREMENT,
debt_id INT,
date DATE NOT NULL,
amount DECIMAL(19,2) NOT NULL,
currency VARCHAR(10)
);
CREATE TABLE debt_status_logs (
id INT PRIMARY KEY AUTO_INCREMENT,
debt_id INT,
total_amount DECIMAL(19,2),
outstanding DECIMAL(19,2),
status_date DATE
);
CREATE TABLE debts (
id INT PRIMARY KEY AUTO_INCREMENT,
source VARCHAR(50) NOT NULL,
total_amount DECIMAL(19,2) NOT NULL,
outstanding DECIMAL(19,2) NOT NULL,
interest_rate DECIMAL(5,2),
currency VARCHAR(10)
);
CREATE TABLE events (
id INT PRIMARY KEY AUTO_INCREMENT,
event VARCHAR(255) NOT NULL,
date DATE NOT NULL,
time TIME NOT NULL
);
CREATE TABLE expenses (
id INT PRIMARY KEY AUTO_INCREMENT,
value DECIMAL(10,2),
expense_date DATE,
currency VARCHAR(20) DEFAULT 'INR',
particulars VARCHAR(255),
debt_id INT,
is_debt_repayment TINYINT(1) DEFAULT 0,
is_earmarked_for_debt_repayment TINYINT(1) DEFAULT 0
);
CREATE TABLE goals (
id INT PRIMARY KEY AUTO_INCREMENT,
name VARCHAR(255) NOT NULL,
date DATE NOT NULL
);
CREATE TABLE linkedin_access_codes (
id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
access_code VARCHAR(255),
retrieved_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
CREATE TABLE linkedin_posts (
id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
post LONGTEXT,
post_id VARCHAR(255),
posted_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
note_id INT,
media_asset_urn TEXT
);
CREATE TABLE mysql_databases (
id INT PRIMARY KEY AUTO_INCREMENT,
description VARCHAR(255),
command VARCHAR(255),
password VARCHAR(255)
);
CREATE TABLE notes (
id INT PRIMARY KEY AUTO_INCREMENT,
note LONGTEXT,
is_published TINYINT(1) DEFAULT 0,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
media_url TEXT,
is_organic TINYINT(1) DEFAULT 1,
published_at DATETIME
);
CREATE TABLE passwords (
id INT PRIMARY KEY AUTO_INCREMENT,
service VARCHAR(255),
username VARCHAR(255),
password VARCHAR(255),
comments TEXT
);
CREATE TABLE reasons (
id INT PRIMARY KEY AUTO_INCREMENT,
goal_id INT,
reason VARCHAR(255) NOT NULL
);
CREATE TABLE runs (
id INT PRIMARY KEY AUTO_INCREMENT,
pre_run_weight_lbs FLOAT DEFAULT 0,
post_run_weight_lbs FLOAT DEFAULT 0,
fat_burn_zone_minutes FLOAT DEFAULT 0,
cardio_zone_minutes FLOAT DEFAULT 0,
peak_zone_minutes FLOAT DEFAULT 0,
distance_covered_kms FLOAT DEFAULT 0,
date DATE,
temperature_in_f FLOAT
);
CREATE TABLE spaced_publications (
id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
note_id INT
);
CREATE TABLE timesheets (
id INT PRIMARY KEY AUTO_INCREMENT,
action_id INT,
date DATE
);
CREATE TABLE tweets (
id INT PRIMARY KEY AUTO_INCREMENT,
tweet LONGTEXT,
tweet_id VARCHAR(255),
posted_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
note_id INT,
media_id TEXT
);
CREATE TABLE virtual_machines (
id INT PRIMARY KEY AUTO_INCREMENT,
description VARCHAR(255),
command VARCHAR(255),
password VARCHAR(255)
);
STEP III - CREATE A VIRTUAL ENVIRONMENT AND INSTALL PYTHON PACKAGES VIA PIP
python3 -m venv botvenv
source botvenv/bin/activate
pip install aiohttp aiosignal async-timeout attrs cachetools certifi charset-normalizer frozenlist google-api-core google-auth google-cloud-core google-cloud-storage google-crc32c google-resumable-media googleapis-common-protos idna multidict mysql-connector-python numpy oauthlib pandas plotext protobuf pyasn1 pyasn1-modules python-crontab python-dateutil python-dotenv pytz requests requests-oauthlib rsa setuptools six tabulate termcolor tzdata urllib3 yarl
deactivate
STEP IV - CREATE AN INITIALIZATION FILE
- Inside the product directory create an initiatization file init.sh, and make it executable along with the main.py file
touch init.sh
chmod +x init.sh
chmod +x main.py
- Get the path to the directory
pwd
- Update the init.sh file with the below content. Be sure to replace path_to_project_directory with the actual path
#!/bin/bash
source path_to_project_directory/botvenv/bin/activate
python3 path_to_project_directory/main.py "$@"
- Create a symbolic link for init.sh and name your bot. Be sure to replace path_to_project_directory with the actual path, and rgw with your preferred bot name (mild narcissism compels me to call my bot by my name initials)
sudo ln -s path_to_project_directory/init.sh /usr/local/bin/rgw
STEP V (OPTIONAL FOR SSH KEY LOGINS TO VIRTUAL MACHINES) - PLACE YOUR SSH KEYS IN THE SSH DIRECTORY, AND SET PERMISSIONS FOR SSH KEYS
chmod 600 path_to_the_directory/files/ssh/*
STEP VI - (OPTIONAL FOR TWITTER MODULE) CONFIGURE THE TWITTER MODULE
- Place your twitter token in the tokens directory, and set permissions for the token
chmod 600 path_to_the_directory/files/tokens/*
- Since cronjobs are typically executed on servers with the default time in UTC, you can adjust your preferred executed_at timezone in fn_publish_queued_tweets method of modules/services/cronjobs/functions.py