Reuse release messages in syncs and releases

[bigboateng]

What changed

• require every generated sync to add at least one append-only release-message fragment
• lead sync PR bodies with the exact new fragment contents and keep provenance collapsed below them
• stage generated output before validation so newly projected, previously untracked fragments are detected
• assemble tagged release bodies from every fragment added since the previous tag
• replace generated PR-title notes with the canonical release-level Markdown
• fail closed when fragments are missing, modified, deleted, or absent from a tag

Why

Intelligence Flow now projects canonical Boatstack release messages, but Boatstack's owned control plane still creates generic sync PR bodies and title-only GitHub releases. These workflow changes complete the downstream slice so the exact reviewed message survives from sync through release.

Impact

The next Intelligence Flow sync will open a PR whose body explains what is being released. Future tagged releases combine all new messages in deterministic filename order without listing commits or diffs.

Validation

• parsed both workflow files as YAML
• ran bash -n against every embedded workflow shell block
• applied the patch to a fresh Boatstack checkout
• dry-ran staged sync detection and tag-range selection
• verified the projected fragment matched the Intelligence Flow source byte-for-byte