This project demonstrates a very small customer care system based on the agent workflow described in PRD.md. It uses a PostgreSQL database for storing complaint tickets.
Configuration is handled by customer_care.config.Config. Values can be customised via environment variables or by editing the file directly.
| Environment Variable | Purpose |
|---|---|
CC_EMAIL_USER |
Email account used for sending messages |
CC_EMAIL_PASSWORD |
Password for the above account |
CC_EMAIL_INBOX_ID |
Inbox identifier to monitor |
CC_RECIPIENTS |
JSON list of internal recipient emails |
CC_DB_HOST |
PostgreSQL host |
CC_DB_PORT |
PostgreSQL port |
CC_DB_NAME |
Database name |
CC_DB_USER |
Database user |
CC_DB_PASSWORD |
Password for the database user |
Run the customer_care/init_db.py script to create the tickets table in your PostgreSQL instance:
python -m customer_care.init_dbThe SQL used by the script is also provided in schema.sql if you prefer running it manually.
Install dependencies and run the tests with coverage:
pip install psycopg2-binary pytest pytest-cov
pytest --cov=customer_careThis repository contains a simple demonstration of an automated customer complaint management workflow. The workflow is composed of modular agents as described in agents.md.
The main.py script runs a mock workflow that fetches a sample complaint email, extracts data, acknowledges the customer, creates a ticket and forwards it to a support team.
python3 main.pyThis is a minimal prototype intended for demonstration purposes only.