Configs that reproduce are erroneously stating that they need a `!bump major` because no config exists

[CodeGat]

See https://github.com/ACCESS-NRI/access-esm1.5-configs/actions/runs/10609553828?pr=86

There are two main issues with the running of this workflow when a config reproduces:

• The pipeline allows merging despite no !bump command being issued - see this annotation on the workflow run.
• The pipeline adds the Failed Repro comment (No config exists) when we are clearly comparing the *-1.0 config, which requires a !bump major despite the repro succeeding.