Skip to content

Setting-up your dev environment

Development environment

If you want to make changes to the code, it is recommended you use our pre-configured development container, which contains all the dependencies you need to run and use blindai.

Firstly, in order to be able to run tests with the server, you'll need to work on an SGX2-ready machine. To find out more about how to do this, check out our installation page.

Now you're ready to set up your development environment!

  • If you are working on the Azure DCsv3 VM, click here**
  • If you are working on your local machine, click here here**

Standard dev environment

To set up our pre-configured development container, you can follow these instructions:

  1. Clone blindai github repo and submodules. 

    git clone https://github.com/mithril-security/blindai --recursive
cd blindai

  2. Make sure you have docker installed on your machine.

  3. If you need to install Docker, you can follow the official Docker installation instructions.

You also need to make sure you haver the correct permissions to run docker commands without sudo. To check this, try running docker run hello-world. If this works, you can skip straight to the next step. If it doesn't, you need to add yourself to docker group: 

sudo usermod -aG docker $USER && newgrp docker
3. Open the blindai folder in VSCode.

  1. Make sure you have the remote container VSCode extension installed. If you don't, install this from the VSCode extensions marketplace.

  2. Open the green menu at the bottom-left of Visual Studio Code. menu

Choose: Dev Containers: Reopen in Container. reopen in container option

This may take some time since there are several dependencies that must be installed.

If you have any issues with this process, make sure you have the BlindAI folder open in your VSCode window.

Building client from source

To compile the client code locally: 

cd client
poetry install

Building server from source

You can build and run the server from source using the justfile: 

just run

Make sure you are in the root of the blindai directory to make use of the justfile commands.

Note that by default the port opened in 9923 is running on http only. For production, we strongly recommend setting up a reverse-proxy that will manage and encrypt the traffic from the client to the blindAI server. Many free reverse-proxy implementations exist, such as caddy, Nginx and Apache:

If you do not set up a reverse proxy, users will need to set the hazmat_http_on_untrusted_port option to True when using blindai's connect() function. Again, this is not recommended for production.

Note that if you make any changes to the server code, a new manifest.toml file will be created when you build the server. In order to be able to connect with the server instance using the BlindAI Core connect() method, you will need to supply a path to a copy of this file in the hazmat_manifest_path option. The manifest.toml files are used during the verification step of the connection progress to check that the server is not running any unexpected and potentially malicious code. You can learn more about this verification process here.

Azure dev environment

To set up our pre-configured development container for your Azure VM, you can follow these instructions:

  1. Clone blindai github repo and submodules. 

    git clone https://github.com/mithril-security/blindai --recursive
cd blindai

  2. Make sure you have docker installed on your machine.

  3. If you need to install Docker, you can follow the official Docker installation instructions.

You also need to make sure you haver the correct permissions to run docker commands without sudo. To check this, try running docker run hello-world. If this works, you can skip straight to the next step. If it doesn't, you need to add yourself to docker group: 

sudo usermod -aG docker $USER && newgrp docker
3. Open the blindai folder in VSCode.

  1. Make sure you have the remote container VSCode extension installed. If you don't, install this from the VSCode extensions marketplace.

  2. Make sure you are connected to your VM as host. To do this, open the green menu at the bottom-left of Visual Studio Code. menu

Choose: Connect to host and select your host. connect to host

  1. Next, open the green menu at the bottom-left of Visual Studio Code again and choose: Dev Containers: Reopen in Container. reopen in container

This may take some time since there are several dependencies that must be installed.

If you have any issues with this process, make sure you have the BlindAI folder open in your VSCode window.

Building client from source

To compile the client code locally: 

cd client
poetry install

Building server from source

You can build and run the server from source using the justfile: 

just run

Make sure you are in the root of the blindai directory to make use of the justfile commands.

Note that by default the port opened in 9923 is running on http only. For production, we strongly recommend setting up a reverse-proxy that will manage and encrypt the traffic from the client to the blindAI server. Many free reverse-proxy implementations exist, such as caddy, Nginx and Apache:

If you do not set up a reverse proxy, users will need to set the hazmat_http_on_untrusted_port option to True when using blindai's connect() function. Again, this is not recommended for production.

Note that if you make any changes to the server code, a new manifest.toml file will be created when you build the server. In order to be able to connect with this server instance using the BlindAI Core connect() method, you will need to supply a path to a copy of this file in the hazmat_manifest_path option. The manifest.toml files are used during the verification step of the connection progress to check that the server is not running any unexpected and potentially malicious code. You can learn more about this verification process here.