How to detect allow_failure exit_codes from custom executor?

GitLab CI/CD
,
1

I am implementing my own gitlab-runner custom executor with Nomad as a backend. Consider the following .gitlab-ci.yml

test:
   script:
       - exit 105
    allow_failure:
        exit_codes:
            - 105

The documentation states that in The Custom executor | GitLab

GitLab Runner provides BUILD_FAILURE_EXIT_CODE environment variable which should be used by the executable as an exit code to inform GitLab Runner that there is a failure on the users job. If the executable exits with the code from BUILD_FAILURE_EXIT_CODE, the build is marked as a failure appropriately in GitLab CI.

This is confusing. How should the detection of “105” go?

Should the execution of the stage return BUILD_FAILURE_EXIT_CODE? Or should the execution of the build_stage ignore documentation and return 105 exit code? Both cases do not work (BUILD_FAILURE_EXIT_CODE is set by gitlab to 1):

$ exit 105
Cleaning up project directory and file based variables   00:00
ERROR: Job failed: exit status 1 

$ exit 105
Cleaning up project directory and file based variables 00:00
ERROR: Job failed (system failure): unknown Custom executor executable exit code 105; executable execution terminated with: exit status 105

The comparison of the actual exit code against the list of exit_codes could be done inside the executor. However, the exit_codes are not passed - I do not see any CUSTOM_ENV_ or other environment variable with it. Where is it stored? Is there a way to extract it? Is this a missing feature and I should make an issue upstream?