.planning/phases/03-spec-10-smoke-verification-dom-event-log-verification/.plan-phase-preferences.md

@@ -0,0 +1,45 @@
+---
+context: phase-3-plan-preferences
+captured: 2026-05-20
+expires_when: /gsd-plan-phase 3 completes
+---
+
+# Phase 3 plan-phase preferences (captured pre-UI-SPEC exit)
+
+User invoked `/gsd-plan-phase 3` 2026-05-20 and answered both initial gate
+questions before workflow exited at UI-SPEC gate. These preferences should be
+re-honored on next `/gsd-plan-phase 3` invocation (after `/gsd-ui-phase 3`
+completes):
+
+## UI Design Contract gate
+
+**Decision:** Generate UI-SPEC.md first → run `/gsd-ui-phase 3` before
+re-running plan-phase.
+
+**Caveat to revisit:** Phase 3 is verification-only — the "page" keyword that
+tripped the gate refers to Puppeteer test fixture pages (synthetic HTML for
+rrweb DOM probing), not user-facing UI. /gsd-ui-phase may produce a thin
+UI-SPEC.md or recommend skipping. Either is acceptable. If it surfaces an
+"is this a UI phase?" question, answer "no — internal test fixtures only".
+
+## Research gate
+
+**Decision:** Research first (light) → spawn gsd-phase-researcher before
+planning.
+
+**Scope for researcher:** light investigation of
+- `puppeteer.Page.metrics()` API for optional RAM scaffolding (D-P3-04)
+- Any new rrweb 2.0.0-alpha.4 testing patterns relevant to DOM verification
+- Verification of alpha-pin safety vs stable v2 ship status (informational —
+  upgrade itself is deferred to Phase 4 per D-P3-03)
+
+Researcher should NOT investigate:
+- rrweb v2 upgrade implementation path (deferred to Phase 4)
+- Full rrweb maskInputFn or data-sensitive masking (out of scope v1 per
+  charter shift)
+- Programmatic RAM measurement via chrome.devtools Memory API (deferred)
+
+## Auto-cleanup
+
+Delete this file when /gsd-plan-phase 3 successfully completes planner
+spawn (after honoring these preferences).