feat: add SmartWardrobe with save_silhouette and get_mix_and_match

[Copilot]

Implements the two missing Pilot features — silhouette persistence (Punto 4) and style-aware mix-and-match suggestions (Punto 3) — as a SmartWardrobe class backed by a BodyAnchors dataclass and a PilotEngine with a PILOT_COLLECTION.

New: api/smart_wardrobe.py

• BodyAnchors — dataclass holding shoulder_w / torso_h scan output.
• PilotEngine — carries PILOT_COLLECTION (5 garments, eg0–eg4) each with brand, name, and recoveryPct.
• SmartWardrobe — core class:
  • save_silhouette(anchors) stores shoulderW, torsoH, timestamp and returns the protection confirmation string.
  • get_mix_and_match(engine, current_look_id) excludes the current look and sorts candidates by nearest recoveryPct before returning the top 4 — correctly implementing the "similar style" intent stated in the spec.

sw = SmartWardrobe("user_001")
sw.save_silhouette(BodyAnchors(shoulder_w=42.5, torso_h=55.0))
# → "Silueta protegida bajo cifrado local."

engine = PilotEngine()
sw.get_mix_and_match(engine, "eg0")
# → ["eg2", "eg3", "eg4", "eg1"]  (sorted by |recoveryPct - eg0.recoveryPct|)

New: tests/test_smart_wardrobe.py

• 16 unit tests covering init state, silhouette overwrite, timestamp validity, recoveryPct-sorted ordering, max-4 cap, and PILOT_COLLECTION schema.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
tryonyou-pilot	Error		Apr 7, 2026 2:10pm