--dry-run shows what a release would change, but it stops short of the steps that can actually fail: writing the files, committing, tagging, running the hooks around them. ferrflow shadow-release runs those steps for real, in a temporary clone of your repository, so a broken hook or a file that will not bump shows up before the real release.

ferrflow shadow-release

The clone gets the versioned files, changelogs, release commit, tags and floating tags, and the pre-commit to pre-publish hooks run against it. Nothing is pushed, no forge release is created, and no publisher or post-publish hook runs. The command prints the commits, files and tags it produced, then deletes the clone. Pass --keep to inspect it instead. A failing step exits non-zero with its error code, so a CI job can run it as a gate.