showdown-environment

Description

This is a Python client-side API for interacting with the Pokemon Showdown website. It also contains an environment to train AI agents.

This repository essentially handles 3 major obstacles in using Pokemon Showdown to train AI agents:

1. Connecting to the Pokemon Showdown server to send challenges, choose moves in battle, etc (see showdown_environment/showdown/)
2. Interpreting the protocol and request messages from Pokemon Showdown into turn-by-turn state objects, which AI agents can use to make informed decisions (see showdown_environment/state/)
3. Collecting the Pokedex, Movedex, Itemdex, Abilitydex, and Typedex from Pokemon Showdown, which can be used to interpret elements of the state objects (see showdown_environment/data/)

Please feel free to use/modify the code for your specific needs.

**Disclaimer: Functionality has only been tested in Ubuntu 20.04.

Running example code

1. Clone the repo.
2. Open a Python virtual environment if necessary, and run pip install -r requirements.txt to install dependencies.
3. Run pip install -e . to install the showdown_environment package contained in the repo.
4. Run cp config_example.json config.json.
5. Fill in config.json. Inspect code in example/ to see how config.json is used.
6. Run python3 example/main.py to see the example code in action.

Update data stored in showdown_environment/data/json/

1. Ensure you have node and npm installed.
2. Run the following code:

   cd scripts
   npm i
   cd ..
   node scripts/update_data.js