Learning Service

A service to learn about Olas agents and Open Autonomy.

System requirements

• Python >=3.10
• Tendermint ==0.34.19
• IPFS node ==0.6.0
• Pip
• Poetry
• Docker Engine
• Docker Compose
• Set Docker permissions so you can run containers as non-root user

Run you own agent

Get the code

1. Clone this repo:

   git clone git@github.com:valory-xyz/academy-learning-service.git
   

2. Create the virtual environment:

   cd academy-learning-service
   poetry shell
   poetry install
   

3. Sync packages:

   autonomy packages sync --update-packages
   

Prepare the data

1. Prepare a keys.json file containing wallet address and the private key for each of the four agents.

   autonomy generate-key ethereum -n 4
   

2. Prepare a ethereum_private_key.txt file containing one of the private keys from keys.json. Ensure that there is no newline at the end.

3. Deploy two Safes on Gnosis (it's free) and set your agent addresses as signers. Set the signature threshold to 1 out of 4 for one of them and and to 3 out of 4 for the other. This way we can use the single-signer one for testing without running all the agents, and leave the other safe for running the whole service.

4. Create a Tenderly account and from your dashboard create a fork of Gnosis chain (virtual testnet).

5. From Tenderly, fund your agents and Safe with some xDAI and OLAS (0xcE11e14225575945b8E6Dc0D4F2dD4C570f79d9f).

6. Make a copy of the env file:

   cp sample.env .env
   

7. Fill in the required environment variables in .env. These variables are:

• ALL_PARTICIPANTS: a list of your agent addresses. This will vary depending on whether you are running a single agent (run_agent.sh script) or the whole 4-agent service (run_service.sh)
• GNOSIS_LEDGER_RPC: set it to your Tenderly fork Admin RPC.
• COINGECKO_API_KEY: you will need to get a free Coingecko API key.
• TRANSFER_TARGET_ADDRESS: any random address to send funds to, can be any of the agents for example.
• SAFE_CONTRACT_ADDRESS_SINGLE: the 1 out of 4 agents Safe address.
• SAFE_CONTRACT_ADDRESS: the 3 out of 4 Safe address.

Run a single agent locally

1. Verify that ALL_PARTICIPANTS in .env contains only 1 address.

2. Run the agent:

   bash run_agent.sh
   

Run the service (4 agents) via Docker Compose deployment

1. Verify that ALL_PARTICIPANTS in .env contains 4 address.

2. Check that Docker is running:

   docker
   

3. Run the service:

   bash run_service.sh
   

4. Look at the service logs for one of the agents (on another terminal):

   docker logs -f learningservice_abci_0