feat(fr-gate): promotion-shape guard — only the train promotes

[LukasWodka]

Hardening from today's parallel-promotion collision (backend#1336 indirect-merging the train's #1326): the gate now fails any PR targeting staging/main/master whose head isn't release-train/* or hotfix-backmerge/*, with an instructive error pointing at the train. Escapes: hotfix label (D14a single-repo lane) and skip-fr-gate (existing audited override). The old fashion stops compiling.

🤖 Generated with Claude Code

---

Note

Low Risk
Workflow-only enforcement on promotion PR shape; escapes for train branches, hotfix back-merges, hotfix label, and skip-fr-gate are unchanged in intent.

Overview
Adds a promotion-shape guard to the reusable FR gate so PRs into staging, main, or master must come from sanctioned lanes, not ad-hoc manual promotions.

The new step runs after the existing skip-fr-gate check and fails the job when the head ref is not release-train/* or hotfix-backmerge/* and the PR lacks the hotfix label. The error message directs teams to the release train workflow and documents the emergency overrides (hotfix, skip-fr-gate). A short comment notes that skip-fr-gate also skips this guard.

^{Reviewed by Cursor Bugbot for commit 0e5f5a7. Bugbot is set up for automated code reviews on this repo. Configure here.}

[LukasWodka]

👋 Heads-up — Code review queue is at 33 / 30

Above the WIP limit. The team convention is to review existing PRs before opening new work.

Open PRs currently in Code review (oldest first):

• averaging-service#222 — fix(sklearn): unwrap Pipeline to terminal estimator before merge dispatch (feat(inventory): assert WHICH checks are required, not just that a caller exists (backend#1681) #209) · author: @divyasinghds · no reviewer assigned
• averaging-service#225 — fix(averaging): fail fast + dead-letter on permission-denied weight write (release-train: staging -> main #221) · author: @divyasinghds · no reviewer assigned
• backend#1201 — docs(rfc): Data spaces & federated dataset composition (draft) · author: @LukasWodka · reviewer: @saadqbal
• backend#1217 — docs(rfc): Dataset integrity & score reproducibility (draft) · author: @LukasWodka · reviewer: @saadqbal
• backend#1261 — RFC-BACKEND-0008: board & process reset (decision RFC — for markup) · author: @LukasWodka · reviewer: @saadqbal
• backend#1284 — docs(deploy): backend deploy runbook skeleton (#1263) · author: @LukasWodka · no reviewer assigned
• backend#1293 — fix(heartbeat): tolerate unknown experiment_key in training_info (#1287) · author: @divyasinghds · no reviewer assigned
• backend#1332 — feat(check-model): allow KNN/GaussianProcess on a single-edge owned dataset (#923 follow-up) · author: @divyasinghds · no reviewer assigned
• claude-skills#14 — feat(skill): add tracebloc-frontend-app skill · author: @waqaskhanroghani · no reviewer assigned
• claude-skills#15 — feat(skill): add tracebloc-design-system skill · author: @waqaskhanroghani · no reviewer assigned

Pull from review before opening new work. (This is a nudge from the kanban WIP check, not a block.)