This is a cooperative board game in which there are two different roles: navigator and party. The navigator role is played by a human, while the cooperating party makes decisions based on LLM responses.
TODO
The process of building a cartesi machine and firing up the co-processor infrastructure is performed through a makefile. In order to fire up the co-processor infrastructure you should use the command
make up
After that you should build the Cartesi Machine used by the co-processor operator. In order to do that, use the command
make build
Next, it's time to make the solver aquire the Cartesi Machine image. The process consists of uploading the Cartesi Machine image to a local AWS S3 like instance and notifying the solver to download it. This is performed with the command
make publish
As time process takes a while (and it's asynchronous) you should check it's status with the command
make publish-status
And wait until it shows the DAG import was completed. Next it's time to tell the Co-processor operator to download the image. To perform this, use the command
make ensure-publish
Now the Co-processor is aware of your image and it's time to deploy the contract that issues tasks and receives callbacks from it. Use the command
make deploy-contract
This step will output the contract address. Now let's interact with the deployed Co-processor. You need to specify the contract address that was just shown as well as the payload. Use the command
CONTRACT_ADDRESS=<ADDRESS>
PAYLOAD=<PAYLOAD>
make send CONTRACT_ADDRESS=$CONTRACT_ADDRESS PAYLOAD=$PAYLOAD
This will send an input to the Co-processor using cast. You can check the docker container log for the operator to see the status of the input processing and once it's completed you can check the solver log for the callback with the results. You can later recover the results from the events generated by the smart contract using the command
source .env && cast logs --rpc-url $RPC_URL --address $CONTRACT_ADDRESS