Deployments offer a few configurable parameters with certain defaults.

The ref parameter can be any named branch, tag, or SHA. At GitHub Enterprise Server we often deploy branches and verify them
before we merge a pull request.

The environment parameter allows deployments to be issued to different runtime environments. Teams often have
multiple environments for verifying their applications, such as production, staging, and qa. This parameter
makes it easier to track which environments have requested deployments. The default environment is production.

The auto_merge parameter is used to ensure that the requested ref is not behind the repository’s default branch. If
the ref is behind the default branch for the repository, we will attempt to merge it for you. If the merge succeeds,
the API will return a successful merge commit. If merge conflicts prevent the merge from succeeding, the API will
return a failure response.

By default, commit statuses for every submitted context must be in a success
state. The required_contexts parameter allows you to specify a subset of contexts that must be success, or to
specify contexts that have not yet been submitted. You are not required to use commit statuses to deploy. If you do
not require any contexts or create any commit statuses, the deployment will always succeed.

The payload parameter is available for any extra information that a deployment system might need. It is a JSON text
field that will be passed on when a deployment event is dispatched.

The task parameter is used by the deployment system to allow different execution paths. In the web world this might
be deploy:migrations to run schema changes on the system. In the compiled world this could be a flag to compile an
application with debugging enabled.

Users with repo or repo_deployment scopes can create a deployment for a given ref.

Merged branch response

You will see this response when GitHub automatically merges the base branch into the topic branch instead of creating
a deployment. This auto-merge happens when:

• Auto-merge option is enabled in the repository
• Topic branch does not include the latest changes on the base branch, which is master in the response example
• There are no merge conflicts

If there are no new commits in the base branch, a new request to create a deployment should give a successful
response.

Merge conflict response

This error happens when the auto_merge option is enabled and when the default branch (in this case master), can’t
be merged into the branch that’s being deployed (in this case topic-branch), due to merge conflicts.

Failed commit status checks

This error happens when the required_contexts parameter indicates that one or more contexts need to have a success
status for the commit to be deployed, but one or more of the required contexts do not have a state of success.

If the repository only has one deployment, you can delete the deployment regardless of its status. If the repository has more than one deployment, you can only delete inactive deployments. This ensures that repositories with multiple deployments will always have an active deployment. Anyone with repo or repo_deployment scopes can delete a deployment.

To set a deployment as inactive, you must:

• Create a new deployment that is active so that the system has a record of the current state, then delete the previously active deployment.
• Mark the active deployment as inactive by adding any non-successful deployment status.

For more information, see “Create a deployment” and “Create a deployment status.”

Users with pull access can view deployment statuses for a deployment:

Users with push access can create deployment statuses for a given deployment.

GitHub Apps require read & write access to “Deployments” and read-only access to “Repo contents” (for private repos). OAuth Apps require the repo_deployment scope.