Liquidation Searcher

Orderly liquidations TPS scale with number of liquidators, so we implement an open source liquidator bot to make it easier to let other traders join the liquidation party.

WARNING: The bot is not fully tested! Use at Your Own Risk!

This project contains two parts:

• A real-time async event-driven trading framework heavily inspired by paradigmxyz/artemis.
• A liquidation searcher that uses the framework to search for liquidation opportunities.

You can read more details from the docs

Source code structure

/docs                                - document
/src/liquidation_searcher            - source code root of the python code
/src/liquidation_searcher/main.py    - entry point of the liquidation searcher
/src/liquidation_searcher/collectors - collectors implementation
/src/liquidation_searcher/strategies - strategies implementation
/src/liquidation_searcher/executors  - executors implementation
/src/liquidation_searcher/engine     - engine implementation
/conf                                - configuration files of different environments
/deployment                          - reference systemd unit file

Configurations

The configurations contain two parts, you can refer to the example files:

plaintext config: staging.yml

app:
  level: "INFO"
  port: 8088 # health check port
orderly:
  account_id: '' # your account id
  rest_endpoint: 'https://testnet-api-evm.orderly.org'
  ws_public_endpoint: 'wss://testnet-ws-evm.orderly.network/ws/stream/'
  ws_private_endpoint: 'wss://testnet-ws-private-evm.orderly.network/v2/ws/private/stream/'
  max_notional: 1000 # max notional of each liquidation claim, it already includes the leverage, should be greater than min notional of the liquidated type claim
  liquidation_symbols: ['PERP_BTC_USDC', 'PERP_ETH_USDC', 'PERP_NEAR_USDC', 'PERP_WOO_USDC'] # liquidation symbols whitelist

secret config: .envrc.example

export ORDERLY_KEY=
export ORDERLY_SECRET=

Run on local machine

The project uses poetry to manage dependencies and virtualenv.

# Required python >= 3.10, you may use asdf or pyenv to manage multiple python versions
pip install poetry
poetry install # install all dependencies in the virtualenv
poetry shell # spawn a shell within the virtual environment
# run it in company's vpn in qa environment or use the staging environment
# source secret envs from your .env file
python src/liquidation_searcher/main.py -c conf/staging.yml

Run with systemd

Reference the systemd unit file liquidation_searcher_dev.service

Run with docker or k8s

Reference the Dockerfile, Note that you need to set the environment variables in the dockerfile or pass them in when running the container in k8s