Chatsky Project Template

This repository is a template for Chatsky projects.

The bot presented here is fully-functional with a very simple script that should be overwritten with yours (more on that in the next section).

This bot uses SQLite to store dialog history. Refer to the context storage tutorials if you'd like to use a different DB engine.

This bot uses Telegram as an interface. Refer to the interface tutorials if you'd like to use different means of communicating.

How to use

[Step 0] System requirements

1. A unix system (e.g. Ubuntu 22.04) is recommended. Although both Chatsky and this template support other systems such as Windows, the instructions below are for unix systems.
2. Git to clone this repository.
3. |Option 1| Docker Engine and Docker Compose if you want to run the bot inside a docker container.
4. |Option 2| Python 3.9+ if you don't want to use docker.

[Step 1] Clone the repo

git clone https://github.com/deeppavlov/chatsky-template.git
cd chatsky-template

Alternatively, you can create a new repository using this one as a template via GitHub:

https://github.com/new?template_name=chatsky-template&template_owner=deeppavlov

[Step 2] Edit project files

1. Write your custom functions (e.g. custom responses, custom services) into services/bot/bot/custom.
2. Edit Chatsky script at services/bot/bot/script.py.
3. Save Telegram bot token into secrets/tg_token.txt.

[Step 3] Test the bot

With Python

First, move to the bot directory:

cd services/bot

Next, install project dependencies:

pip install -r requirements.txt

Finally, run tests:

pytest test.py

With Docker

docker build --target test services/bot

[Step 4] Run the bot

With Python

First, you need to be in the root directory of the project.

If you've done the previous step, run

cd ../..

to return from the services/bot directory.

Next, install project dependencies (if you haven't done so yet):

(cd services/bot && pip install -r requirements.txt)

Deploy the bot to telegram with:

(export DB_URI="sqlite+aiosqlite:////$(pwd)/sqlite.db" TG_BOT_TOKEN_FILE="$(pwd)/secrets/tg_token.txt" && cd services/bot && python app.py)

With Docker Compose

docker compose up