CREAMbots is the collection of python tools to work with liquidity pools on EVM chains. The actual bots will continue to grow and as of the initial commit, are simply a proof of concept. There are however standardized fetchers and arbitrage path builders available for data collection and organization. Much of this work has been inspired by BowTiedDevil. Some of the LP/Liquidity/Arb path builders are directly based on his work. Go check out his stack Degen Code for great insight into blockchain work with Python and Vyper. TY Devil!
- Python version 3.10 or newer.
- Redis (website)
- degenbot (pypi | github): It should be auto installed when you install this and will install it's own set of dependencies, which this app relies on in turn.
- eth-ape (pypi | github): It should be auto installed when you install this.
- networkx (pypi): Used for three-pool arbitrage path building. It should be auto installed when you install this.
- redis (pypi): Used to interact with a redis server. It should be auto installed when you install this.
- tqdm (pypi): Used for progress meters around the app. It should be auto installed when you install this.
- ujson (pypi): Used to parse JSON. It should be auto installed when you install this.
- CREAM (github): If you want to actually react to blockchain transactions/events, you need to install this and run it in a separate process. This isn't a package yet, so you need to
git cloneit and install it as an editable installation in a separate folder. - CREAMchains (github): You'll need this installed to have access to the chain data (things like factories, routers, rpcs, etc). This isn't a package yet, so you need to
git cloneit and install it as an editable installation in a separate folder.
At the moment the only way to install is from source. Use git clone to create a local copy of this repo, then install with pip install -e /path/to/repo. This creates an editable installation that can be imported into a script or Python REPL using import cream_bots.
You'll need to do a bit of legwork to get your environment set up. Once that is set there are a handful of helper scripts that you can run to keep your chain data up to date for use in bots/apps.
At the moment the supported chains are:
- arbitrum (alchemy / local node)
- avalanche (infura / local node)
- base (alchemy / local node)
- ethereum (alchemy / local node)
- optimism (alchemy / local node)
- polygon (alchemy / local node)
Once you have a redis server running, this tool will subscribe to messages to these channels
cream_eventscream_pending_transactionscream_finalized_transactions
The main CREAM app publishes to these channels. Depending on the chain, either pending or finalized transactions channels are used. Base and Optimism don't have pending transactions so you can only see them after they are confirmed in a block. The rest should work with pending transactions. Arbitrum uses the sequencer. This will certainly be updated during development.
The app expects Redis to be local on port 6379 when you run things. You can alter the host/port as needed in config/constants.py.
You'll need to add a few things to your .bashrc/.zshrc to ensure the connections can be made. I highly recommend using Alchemy if you don't have a local node. If you do, just configure things for that. See the shell-example.txt file for how to add those. The other CREAM tools rely on Ape for a lot of things so you'll see some ape-specific stuff in various files. The builders don't require Ape, but forthcoming bots will expect that you are managing your accounts with it so you'll need it installed and configured for the chains you are going to use.
There are a variety of scripts that you can run to pull data from chain and find arbitrage pathways. They are all located in /builders/ They write json files to /data/ by chain. This will take a LONG time if you are using a public RPC. To that end. I've published a snapshot and will try to keep it updated. You can download a version from 2024-09-18 here: amazon s3. Extract the files into /data/ and you're good to go. Note: the Arbitrum liquidity data is from 2024-03-22.
The fetchers to retrieve Liquidity Pool (LP) data from chain factories is in /builders/. You call the LP fetchers like so:
cream_lps_v2 gets all V2 pools on all chains
cream_lps_v2 ethereum gets all V2 pools only on ethereum
cream_lps_v3 gets all V3 pools on all chains
cream_lps_v3 ethereum gets all V3 pools only on ethereum
After you have V3 pools fetched, you can get liquidity data for them. The fetcher to retrieve V3 liquidity data is in /builders/. You call the liquidity fetcher like so:
cream_liquidity gets liquidity on all chains
cream_liquidity ethereum gets liquidity only on ethereum
After you have your LP data, you can create two- and three-pool arbitrage pathways. The builders for these are in /builders/. You call the arbitrage pathwy builders like so:
cream_arbs_2pool gets all two-pool arbitrage pathways on all chains
cream_arbs_2pool ethereum gets all two-pool arbitrage pathways only on ethereum
cream_arbs_3pool gets all three-pool arbitrage pathways on all chains
cream_arbs_3pool ethereum gets all three-pool arbitrage pathways only on ethereum
That's about it for this module. Once you've pulled all the data, import this module into your bots to bootstrap pools/liquidity/arb paths etc.
This is undoubtedly busted in several ways, but I'll be working on it for personal use so it should get refined.