docs(catalogue): add play-safe spicy draft

docs/product/question-drafts/play-safe-spicy-v1.md

@@ -0,0 +1,34 @@
+# Play-safe spicy question draft v1
+
+Status: source-only, unpublished, not approved, imported, or published. This is an editorial draft for human policy review.
+
+## Packs
+
+- private-signals: 92 questions.
+- spark-and-slow-burn: 92 questions.
+- lead-and-follow: 91 questions.
+- shared-curiosity: 91 questions.
+- after-dark-conversations: 92 questions.
+- private-possibilities: 92 questions.
+
+## Exact distribution
+
+- 550 total: 300 swipe, 100 text_answer, 70 audio, 40 photo, 40 who_likely.
+- Intensities: 120 at level 2, 300 at level 3, 130 at level 4.
+- Swipe: 120 asymmetric concepts as 120 primaries plus 120 inverse rows, and 60 symmetric rows.
+
+## Current in-app outcomes
+
+- Swipe proposals create a mutual yes or maybe result. Asymmetric rows use reversed partner wording and inverse links.
+- Text answers let each partner provide the final private detail in their own words.
+- Audio prompts create a private 60-second spoken result.
+- Photo prompts request only safe clothed outfits, objects, notes, lighting, and settings.
+- Who-is-more-likely prompts create a light comparison and a natural private follow-up.
+
+## Policy boundaries
+
+Developer-authored wording avoids named sexual acts, anatomy, sex toys, fetish labels, nudity, public sexual activity, coercion, and instructional wording. Intensity 4 uses edge-reviewed language about adult private dynamics, boundaries, roles, surprise, and user-led detail. This is not a claim of Google approval.
+
+Known limitation: the current API accepts response_data loosely. This source draft assumes no structured compatibility builder or safe user-generated-content moderation envelope, and uses only existing supported question types.
+
+Run node scripts/validate-question-draft.mjs to verify this source-only contract.