In what situation you find that your jobs are rerun needlessly? I can think of several possibilities: commit, retrying failed pipeline, manually triggered pipeline...

This may be different for each team, but in my case, specifically regarding applications and services, each commit represents a significant piece of work and serves as a release candidate. When these commits are finally merged into the main branch, each will be treated as a potential release.

That being said, to optimize this process, I would like to prevent the re-execution of jobs that have already run successfully for any given pipeline because the results are idempotent, meaning that running them again will not produce different outcomes — when constrained to the context of a given pipeline.

So basically I'm focused on preventing the re-execution of jobs that have succeeded; I understand that jobs can fail for various reasons, but re-running a failed job often doesn’t make sense unless the failure was sporadic, such as due to an agent failure.

More importantly, I would like to prevent deployments from previous releases.

Right so you mean following situation?

1. A commit is pushed to branch A and a pipeline is triggered.
2. The branch gets merged to main and the pipeline is triggered again.
3. (optionally) a tag gets pushed to the repository, triggering yet another pipeline.

That's correct — those are the workflows. However, the third option is not allowed because it overlaps with the second one.

But the question is more general, and not tied to any workflow in particular. I'm just looking for a built-in option to disable or prevent a specific job from running again after it has successfully completed in the pipeline.