Game development API that allows interaction with the Aleo Zero-Knowledge platform.

Check Boloney! the first game built with ZK Gaming Toolkit.

There are 2 main ways to run the API: locally or inside an isolated Minikube container.

Before running the API locally, make sure to install the following software:

• Node.js version 18.16.0 LTS
• Yarn
• Aleo
• Leo
• SnarkOS
• Aleo development server - run cargo install aleo-development-server

Run a local SnarkOS beacon node:

snarkos start --nodisplay --dev 0 --beacon "APrivateKey1zkp8CZNn3yeCseEtxuVPbDCwSyhGW6yZKUYKfgXmcpoGPWH"

⚠️ Do not change the private key, since the app is configured to use that in develop.

Open another terminal window and run Aleo development server:

aleo-develop start -p http://127.0.0.1:3030

⚠️ Make sure to specify that local address. If no address is specified, the dev server will connect to the public testnet and you usually don't want that when developing.

Build all the programs locally by running:

./build_local_programs.sh

The first time you run the API, make sure to deploy the programs as well:

DEPLOY_PROGRAMS=true DEPLOY_PRIVATE_KEY=APrivateKey1zkp8CZNn3yeCseEtxuVPbDCwSyhGW6yZKUYKfgXmcpoGPWH yarn start

Pay very close attention to the application and development server logs.

After all the programs have been deployed to the network, the API should be accessible at http://localhost:5001

Unless you reset the network, you don't need to re-deploy the programs, so the following time you want to run the API locally, just run:

yarn start

If you wish to reset your network, stop the beacon process and then run:

snarkos clean --dev 0

⚠️ After you reset the local network, you will have to re-deploy your programs.

⚠️ It is highly discouraged to use this method unless you are working on deployment or you know what you are doing. When you are running the API through minikube, the app will connect to the public testnet and that is not ideal for development.

Before running the API with Minikube, make sure to install the following software:

• Docker
• kubectl
• Minikube - perform only the first step, titled "Installation"
• Skaffold

Before running Minikube, make sure Docker is running. If you are on MacOS, make sure to give Docker access to at least 8GB of memory and at least 50GB of virtual disk. You can do that by opening Docker desktop -> settings (in the top right corner) -> resources.

Run Minikube:

minikube start --cpus=max --memory=max

Enable ingress add-on:

minikube addons enable ingress

Open a new terminal tab and run:

sudo minikube tunnel

This will allow you to access the deployed application at the address specified in the ingress configuration so keep it running in the background.

Return to the first terminal tab and run:

skaffold run

This will build the API and deploy it on the minikube cluster.

To check the status of your pods, run:

kubectl -n zk-gaming-tk-local get pods

To read the logs, run:

# pod_name can be retrieved from the output of the previous command
kubectl -n zk-gaming-tk-local logs <pod_name> -f

If the pods are running correctly, the API should be accessible at http://zk-gaming-tk.localhost

⚠️ On MacOS you may need to configure dnsmasq in order to access custom domain names. Consider following this guide and use .localhost instead of .test and .box.

⚠️ These steps can only be completed by maintainers that have access to the wallet.

If a PR contains any changes to the Leo programs, make sure to perform the following actions before completing it so that the programs are correctly deployed to the testnet:

1. Navigate here and give some credits to the Kryha Aleo account.
2. Wait for the transaction to complete. Don't close the browser tab!
3. Complete the PR and the pipeline should automatically deploy the new programs to the testnet using the newly obtained credits. Just be patient... Like, very patient... Deployment through the pipeline takes about 8 minutes per program.

When creating a new Leo program, the pipeline has to be updated in order for the program to be deployed to the testnet.

Let's assume the name of our program is cool_program. For it to be properly deployed, open azure-pipelines.yaml and add the following case to the run_check_script job:

files=$(git diff HEAD HEAD~ --name-only)

while IFS= read -r name; do
    # previous cases are omitted here
    ...
    elif [[ $name =~ ^contracts/cool_program/* ]]; then
        echo "##vso[task.setvariable variable=coolProgramUpdated;isoutput=true]True"
    fi
done <<<"$files"

Then, in build_and_deploy_leo_programs_stage define a new variable:

variables:
    # previous variables omitted
    ...
    coolProgramUpdated: $[stageDependencies.check_updated_programs.run_check_script.outputs['UpdatedPrograms.coolProgramUpdated']]

In build_and_deploy_leo_programs_job job, after all the previously defined steps, add these new steps:

- script: |
    docker build -f Dockerfile.program . \
      --build-arg APP_NAME=cool_program \
      --build-arg PRIVATE_KEY=$(privateKey) \
      --build-arg BUILD_ID=$(buildId) \
      --build-arg FEE=600000 \
      --build-arg ZK_GAMING_ALEO="eu.gcr.io/web3-335312/aleo/zk-gaming-snarkos:latest"
  displayName: Cool Program Docker build
  condition: and(succeeded(), eq(variables['coolProgramUpdated'], 'True'))
  retryCountOnTaskFailure: 3
- script: echo "##vso[task.setvariable variable=coolProgramVersion;isoutput=true]$(buildId)"
  displayName: Update Cool Program version locally
  condition: and(succeeded(), eq(variables['coolProgramUpdated'], 'True'))
- script: |
    az pipelines variable-group variable update \
      --group-id $(aleoProgramIdsGroupId) \
      --name coolProgramVersion \
      --org $(devOpsOrg) \
      --project $(devOpsProject) \
      --value $(buildId)
  displayName: Update Cool Program version in variable group
  condition: and(succeeded(), eq(variables['coolProgramUpdated'], 'True'))
  env:
    SYSTEM_ACCESSTOKEN: $(System.AccessToken)

The version variable has to be added to the pipeline variable group. For that, contact [email protected].

Since both Leo and SnarkOS are currently in active development and some changes may break the build, we manually set the commit hash in Dockerfile.snarkos to the latest versions that work properly with the toolkit. If you wish to update to a newer version you should:

1. git pull the newest Leo or SnarkOS code and run cargo install --path . to install the new versions locally.
2. Make sure the programs build and run correctly with the updated CLI tools.
3. Run the toolkit locally with yarn start and make sure the endpoints work as expected.
4. Run git log in the Leo or SnarkOS repo and copy the HEAD commit hash.
5. Open Dockerfile.snarkos and set the proper commit hash to the pasted value.
6. Run the toolkit through skaffold run and make sure it works properly.