The default strategy, relies on pull requests into the main branch, the pipelines will then automatically push out a new release branch for the impacted repository and any dependencies.
gitGraph
commit
branch issues/1-example order: 1
commit
switch main
merge issues/1-example
commit id: "PR issue 1 to main"
branch release/0.1.0 order: 3
checkout release/0.1.0
commit id: "Release 0.1.0"
switch main
commit
branch issues/2-example order: 2
commit
commit
commit
switch main
merge issues/2-example
commit id: "PR issue 2 to main"
branch release/0.1.1 order: 4
checkout release/0.1.1
commit id: "Release 0.1.1"
switch release/0.1.0
branch hotfix/3-example order: 5
checkout hotfix/3-example
commit
switch release/0.1.0
merge hotfix/3-example
commit id: "PR to release branch"
switch main
merge release/0.1.0
commit id: "PR from release to main"
branch release/0.1.2 order: 6
checkout release/0.1.2
commit id: "Release 0.1.2"
The downside of this strategy is that it will create a large number of branches on each repository, however this is on purpose as it will automatically update the various flux configurations, helm charts, image versions, etc making it easier to manage the large number of repositories overall.