Variables and secrets

Reference variables in your pipeline

Variables are configured as environment variables in the build container. You can access the variables from the bitbucket-pipelines.yml file or any script that you invoke by referring to them in the following way: 

where AWS_SECRET is the name of the variable.

Default variables

Pipelines provides a set of default variables that are available for builds, and can be used in scripts.

You can override the default variables by specifying a variable with the same name.

User-defined variables

You can add, edit, or remove variables at the workspace, repository, and deployment environment levels. If you use the same name as an existing variable, you can override it. The order of overrides is Deployment > Repository > Workspace > Default variables. Each deployment environment is independent so you can use the same variable name with different values for each environment.

Before you begin bear in mind that:

• Names can only contain ASCII letters, digits and underscores

• Names are case-sensitive

• Names can't start with a digit

• Variables defined by the shell should not be used. You can find them by using a step with the command printenv.

Workspace variables

Variables specified for a workspace can be accessed from all repositories that belong to the workspace. You must be an administrator to manage workspace variables.

1. From your profile. avatar, select a workspace.

2. Select the Settings cog on the top navigation bar.

3. Select Workspace settings from the Settings dropdown menu.

4. In the menu on the left, go to  Pipelines > Workspace variables.

• Workspaces variables can be overridden by repository variables.

• Workspace variables can be accessed by all users with the write permission for any repository (private or public) that belongs to the team or account.

• You must be an administrator of a workspace or a repository to manage variables respectively.

Repository variables

Pipelines variables added at the repository level can be used by any user who has write access in the repository. To access and configure the repository variables, the user must be an admin of that repository.

From the repository, you can manage repository variables in Repository settings > Pipelines > Repository variables. Note: Repository variables override team variables.

Deployment variables

You can also define variables so that they can only be used in a specific deployment environment.

From the repository, you can manage deployment variables in Repository settings > Pipelines > Deployments. Note: Deployment variables override both team and repository variables, and are unique to each environment.

Secured variables

You can secure a variable, which means it can be used in your scripts but its value will be hidden in the build logs (see example below). If you want to edit a secure variable, you can only give it a new value or delete it.  Secure variables are stored as encrypted values. Click the padlock to secure the variable.

Secured variable masking

Pipelines masks secure variables so they are not disclosed to your team members viewing build logs. If a value matching a secured variable appears in the logs, Pipelines will replace it with $VARIABLE_NAME.

This can lead to confusion about whether secured variables are working properly, so here's an example of how it works:

First, we have created a secure variable, MY_HIDDEN_NUMBER, with a value of 5.

Then we used this variable in the YAML file:

The value of the variable can be used by the script, but will not be revealed in the logs. It is replaced with the name of the variable, $MY_HIDDEN_NUMBER.

If you have secure variable value set to a common word, that word will be replaced with the variable name anywhere it appears in the log file. Secured variables are designed to be used for unique authentication tokens and passwords and so are unlikely to be also used in clear text.

Pipelines also matches some basic encodings of the variable value, like URL encoding, to prevent variables being displayed when used in URLs.

Third-party secret providers

If your organization uses a third-party secret management tool such as Hashicorp Vault to store or cycle secrets, Bitbucket Pipelines can integrate with that tool to retrieve secrets from that tool as needed.

To use a third-party secret provider:

1. Code middleware to let your third-party secret provider communicate with Bitbucket.

2. Set up your Bitbucket runner to contact your third-party secret provider.

3. Set up your Bitbucket pipeline to retrieve secrets

1. Code middleware

We’ve provided a sample repository to get you started. This repository contains an example java app for a third-party secret provider integration.

This app:

• Allows the provider to respond to secret requests from Bitbucket,

• Validate the request is coming from Bitbucket using an OIDC JWT token.

• Return a basic set of secrets.

Use this example as a starting point to code your own middleware specific to your third-party provider.

As you code your middleware, make sure to:

• Use the Open API spec we’ve provided in the sample repository to ensure your middleware can communicate with Bitbucket.

• Specify which secrets you want your third-party provider integration to send to Bitbucket. For an example of how to filter secrets, see the file SecretStoreImpl.java in our sample repository.

2. Set up your Bitbucket runner

To retrieve secrets from your third-party provider, you’ll need to:

1. Set up your pipeline’s self-hosted runner to connect to your middleware.

2. Specify which secrets you’d like to retrieve from your provider.

Set up your pipeline’s self-hosted runner

Your self-hosted runner needs to connect to your third-party secret provider at startup. To do this, you’ll need to provide the URI of the middleware you’ve written for your provider. Here’s how to do this for various kinds of self-hosted runner:

Docker

Set SECRET_PROVIDER_URI environment variable when starting the container:

Shell

Add --secretProviderUri property when calling start.sh script:

Powershell

Add -secretProviderUri property when calling start.ps1 script:

3. Set up your Bitbucket pipeline

To retrieve a secret from your third-party secret provider:

• Enable OICD for each pipeline step that will retrieve a secret from your provider.

• Use a variable that matches the name of a secret stored on your provider.

Your pipeline will automatically replace the current value of that variable with the secret hosted on your provider.

For example, if your pipeline contains a variable called $SECRET with a value of 1234 and your third-party secret provider contains a corresponding secret with the same name, your pipeline will replace 1234 with the appropriate secret value at runtime.