Improve CI pipeline

LHCb does: https://gitlab.cern.ch/lhcb-dirac/LHCbDIRAC/-/blob/master/.gitlab-ci.yml#L304-324

TODO:

• Change CI stages order:
  • deploy the version on Pypi (or deploy a pre-release version)
  • update the majestix version with this new version
  • run the tests
• add systems tests in the test report
• use the WMS kind deployment to run some preliminary tests (optional): https://gitlab.cta-observatory.org/cta-computing/dpps/workload/wms/-/blob/main/Makefile?ref_type=heads
• use the harbor proxy_cache for the Docker images, ex: image: harbor.cta-observatory.org/proxy_cache/python:3.12

• !246 (merged)
  • use the dpps-aiv-toolkit in the ci (waiting for cta-computing/dpps/aiv/dpps-aiv-toolkit!21 (merged))
  • fill the traceability matrix with UCs and Reqs
• !228 (merged)
  • Run pre-commit, on commited files only at first, since old ones do not respect pep syntax
  • Generate test report: https://gitlab.cta-observatory.org/cta-computing/dpps/workload/notes/test-report/-/issues/4
• !231 (merged)
  • Store Docker images Harbor