Skip to main content
Interact with various DeFi protocols. Non-custodial.Tools return unsigned transactions. The toolkit is built on top of a Universal DeFi API (Compass API) allowing agents to perform financial operations like:
  • Swapping tokens on Uniswap and Aerodrome
  • Lending or borrowing assets using protocols on Aave
  • Providing liquidity on Aerodrome and Uniswap
  • Transferring funds between wallets.
  • Querying balances, portfolios and monitoring positions.

​
Overview

​
Integration details

ClassPackageSerializableJS supportVersion
LangchainCompassToolkitlangchain-compass❌❌PyPI - Version

​
Tool features

Here’s a sample of the tools this toolkit provides (subject to change daily):
  • aave_supply: Supply assets to Aave to earn interest.
  • aave_borrow: Borrow assets from Aave using collateral.
  • uniswap_swap_sell_exactly: Swap a specific amount of one token on Uniswap.
  • generic_portfolio_get: Retrieve a wallet’s portfolio in USD and token balances.
  • generic_transfer_erc20: Transfer ERC20 tokens between addresses.

​
Setup

Here we will:
  1. Install the langchain package
  2. Import and instantiate the toolkit
  3. Pass the tools to your agent with toolkit.get_tools()

​
Installation

This toolkit lives in the langchain-compass package: 
pip install -qU langchain-compass

​
Environment Setup

To run these examples, ensure LangChain has access to an LLM service. For instance, if you’re using GPT-4o, create a .env file containing: 
# .env file
OPENAI_API_KEY=<your_openai_api_key_here>

​
Instantiation

Now we can instantiate our toolkit: 
from langchain_compass.toolkits import LangchainCompassToolkit

toolkit = LangchainCompassToolkit(compass_api_key=None)

​
Tools

View available tools: 
tools = toolkit.get_tools()
for tool in tools:
    print(tool.name)

# Expected output:

aave_supply
aave_borrow
aave_repay
aave_withdraw
aave_asset_price_get
aave_liquidity_change_get
aave_user_position_summary_get
...

​
Invocation

To invoke a single tool programmatically: 
tool_name = "generic_ens_get"
tool = next(tool for tool in tools if tool.name == tool_name)
tool.invoke({"ens_name": "vitalik.eth", "chain": "ethereum:mainnet"})

EnsNameInfoResponse(wallet_address='0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045', registrant='0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045')

​
Use within an agent

We will need a LLM or chat model: 
from dotenv import load_dotenv
from langchain_openai import ChatOpenAI

load_dotenv()

llm = ChatOpenAI(model="gpt-4o")
Initialize the agent with the tools: 
from langchain.agents import create_agent


tools = toolkit.get_tools()
agent_executor = create_agent(llm, tools)
Example usage: 
example_query = "please set an allowance on Uniswap of 10 WETH for vitalic.eth."  # spelt wrong intentionally

events = agent_executor.stream(
    {"messages": [("user", example_query)]},
    stream_mode="values",
)
for event in events:
    event["messages"][-1].pretty_print()
Expected output: 
================================ Human Message =================================

please set an allowance on Uniswap of 10 WETH for vitalic.eth.
================================== Ai Message ==================================
Tool Calls:
  generic_ens_get (call_MHIXRXxWH0L7iUEYHwvDUdU1)
 Call ID: call_MHIXRXxWH0L7iUEYHwvDUdU1
  Args:
    chain: ethereum:mainnet
    ens_name: vitalic.eth
================================= Tool Message =================================
Name: generic_ens_get

wallet_address='0x44761Ef63FaD902D8f8dC77e559Ab116929881Db' registrant='0x44761Ef63FaD902D8f8dC77e559Ab116929881Db'
================================== Ai Message ==================================
Tool Calls:
  generic_allowance_set (call_IEBftbtBfKCkI1zFXXtEY8tq)
 Call ID: call_IEBftbtBfKCkI1zFXXtEY8tq
  Args:
    amount: 10
    chain: ethereum:mainnet
    contract_name: UniswapV3Router
    sender: 0x44761Ef63FaD902D8f8dC77e559Ab116929881Db
    token: WETH
================================= Tool Message =================================
Name: generic_allowance_set

{"type": "unsigned_transaction", "content": {"chainId": 1, "data": "0x095ea7b300000000000000000000000068b3465833fb72a70ecdf485e0e4c7bd8665fc450000000000000000000000000000000000000000000000008ac7230489e80000", "from": "0x44761Ef63FaD902D8f8dC77e559Ab116929881Db", "gas": 46434, "to": "0xC02aaA39b223FE8D0A0e5C4F27eAD9083C756Cc2", "value": 0, "nonce": 79, "maxFeePerGas": 2265376912, "maxPriorityFeePerGas": 6400594}}

​
API reference

langchain-compass is built on top of the Compass API. Each tool corresponds to an API endpoint. Please check out the docs here
Edit the source of this page on GitHub.
Connect these docs programmatically to Claude, VSCode, and more via MCP for real-time answers.
⌘I