[Change] Unified Policy Engine (DoR/DoD/Flow/PI) — Δ1

[djm81]

Why

Teams love tools that enforce working agreements consistently. Today DoR/DoD are fragmented across features; Kanban/SAFe policies are not first-class. A single Policy framework with policy.validate (hard failures; deterministic) and policy.suggest (AI-assisted; confidence-scored; patch-ready) gives one mechanism for DoR, DoD, Kanban entry/exit, and SAFe PI readiness so refinement, planning, and standup share the same quality gates.

What Changes

• Introduce a single "Policy" framework: policy.validate (hard failures; deterministic), policy.suggest (AI-assisted; confidence-scored; patch-ready).
• First policies: Scrum (DoR + DoD), Kanban (entry/exit per column), SAFe (PI readiness hooks, minimal baseline).
• Machine-readable output: JSON for CI gates, Markdown for humans.
• Config: .specfact/policy.yaml; specfact policy validate runs without network access (against snapshots when applicable).
• Policy results include: rule id, severity, evidence pointer (field/path), recommended action.

Acceptance Criteria

• A project can define policies in .specfact/policy.yaml.
• specfact policy validate runs without network access (against snapshots when applicable).
• Policy results include rule id, severity, evidence pointer, recommended action.

---

OpenSpec Change Proposal: unify-policies-engine