AR-House/scripts/test_wave2_green_cove.py

"""Test focalizado de Wave 2 (ValueEstimator + OfferStrategist) con deal real.

DEAL: 446 VERMONT Avenue, Green Cove Springs, FL 32043 — $219K MLS retail
- Clay County (NO court_records scraper — soft-fail NOT_IMPLEMENTED esperado)
- 2005 build, 3/2, 1461 sqft, tax $3348, insurance $876, rent $1789, no HOA
- ARV $240K (estimacion razonable para zona Class B/C 2005 build)

Criterios validados (8 puntos del spec del usuario):
1. ValueEstimator: comparison listing vs comps
2. ValueEstimator: rango low/mid/high
3. OfferStrategist: Strike/Stretch/Walk-Away
4. OfferStrategist: angulo de ataque (psychological points)
5. OfferStrategist: contra-ofertas anticipadas
6. OfferStrategist: si auction → MAB, si MLS → Strike (este deal es MLS)
7. Briefing: seccion "Valor Real vs Listing"
8. Briefing: seccion "Oferta Recomendada"
"""
from __future__ import annotations

import io
import sys
import time
from pathlib import Path

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
ROOT = Path(__file__).resolve().parent.parent
sys.path.insert(0, str(ROOT))

from orchestrator import DealInputs, BuyerProfile, analyze_deal  # noqa: E402


def status_cb(msg: str) -> None:
    print(f"[{time.strftime('%H:%M:%S')}] {msg}", flush=True)


def main() -> int:
    print("=" * 70)
    print("Wave 2 Validation — 446 Vermont Ave, Green Cove Springs FL $219K MLS")
    print("=" * 70)

    deal = DealInputs(
        address="446 VERMONT Avenue, Green Cove Springs, FL 32043",
        price=219_000,
        rent=1_789,
        property_tax=3_348,
        insurance=876,
        hoa=0,
        sqft=1_461,
        beds=3,
        baths=2.0,
        year_built=2005,
        arv=240_000,       # estimacion razonable; sin rehab significativo para 2005 build
        rehab_override=8_000,  # cosmetics solo, no major rehab
        deal_type="mls",       # NORMAL MLS retail, no auction
    )
    profile = BuyerProfile(
        profile_class="C",
        fico=720,
        capital_available=65_000,
        nationality="Argentina",
    )

    print(f"DEAL: {deal.address}")
    print(f"  price=${deal.price:,} rent=${deal.rent:,}/mo arv=${deal.arv:,}")
    print(f"  beds={deal.beds}/baths={deal.baths} sqft={deal.sqft} year={deal.year_built}")
    print(f"  tax=${deal.property_tax:,}/y insurance=${deal.insurance:,}/y hoa=${deal.hoa}/mo")
    print(f"  deal_type={deal.deal_type}, rehab_override=${deal.rehab_override:,}")
    print()

    t0 = time.perf_counter()
    result = analyze_deal(deal, profile, photo_bytes=None, status_cb=status_cb)
    elapsed = time.perf_counter() - t0

    print()
    print("=" * 70)
    print(f"COMPLETADO en {elapsed:.0f}s ({elapsed/60:.1f} min)")
    print("=" * 70)

    # Get outputs
    ve_out = (result.value_estimate or {}).get("output", "") or ""
    os_out = (result.offer_strategy or {}).get("output", "") or ""
    briefing_out = (result.executive_briefing or {}).get("output", "") or ""

    # ===========================================================
    # 8 CRITERIOS DE VALIDACION (spec del usuario)
    # ===========================================================
    print()
    print("─" * 70)
    print("VALIDACION 8 CRITERIOS")
    print("─" * 70)

    # 1. ValueEstimator: comparison listing vs comps
    c1 = any(kw in ve_out.lower() for kw in [
        "listing", "precio listado", "$219", "comp", "comparable"
    ])
    print(f"  1. ValueEstimator: comparison listing vs comps: {'✅' if c1 else '❌'}")

    # 2. ValueEstimator: rango low/mid/high
    has_low = "low" in ve_out.lower() or "bajo:" in ve_out.lower()
    has_mid = "mid" in ve_out.lower() or "medio:" in ve_out.lower()
    has_high = "high" in ve_out.lower() or "alto:" in ve_out.lower()
    c2 = has_low and has_mid and has_high
    print(f"  2. ValueEstimator: rango low/mid/high: {'✅' if c2 else '⚠️ (parcial)'}")

    # 3. OfferStrategist: Strike/Stretch/Walk-Away
    has_strike = "strike" in os_out.lower()
    has_stretch = "stretch" in os_out.lower()
    has_walkaway = "walk-away" in os_out.lower() or "walk away" in os_out.lower()
    c3 = has_strike and has_stretch and has_walkaway
    print(f"  3. OfferStrategist: Strike/Stretch/Walk-Away: {'✅' if c3 else '⚠️ (parcial)'}")

    # 4. OfferStrategist: angulo de ataque
    c4 = any(kw in os_out.lower() for kw in [
        "angulo de ataque", "ángulo de ataque", "angulo", "psicologic", "psicológ",
        "presentacion", "argumento",
    ])
    print(f"  4. OfferStrategist: angulo de ataque (psicologico): {'✅' if c4 else '❌'}")

    # 5. OfferStrategist: contra-ofertas anticipadas
    c5 = any(kw in os_out.lower() for kw in [
        "contra-oferta", "contra oferta", "counter-offer", "counter offer",
        "contraoferta", "anticipad",
    ])
    print(f"  5. OfferStrategist: contra-ofertas anticipadas: {'✅' if c5 else '❌'}")

    # 6. MLS deal → debe usar Strike NO MAB (este test es MLS)
    has_mab_only = "mab" in os_out.lower() and not has_strike
    c6 = has_strike and not has_mab_only
    print(f"  6. MLS deal usa Strike (no MAB exclusivo): {'✅' if c6 else '❌'}")

    # 7. Briefing: seccion "Valor Real" / "Valor vs Listing"
    c7 = any(kw in briefing_out.lower() for kw in [
        "valor real", "valor estimado", "estimación de valor", "estimacion de valor",
        "valor vs listing", "valor del inmueble",
    ])
    print(f"  7. Briefing incluye seccion Valor Real: {'✅' if c7 else '❌'}")

    # 8. Briefing: seccion "Oferta Recomendada"
    c8 = any(kw in briefing_out.lower() for kw in [
        "oferta recomendada", "recomendación de oferta", "recomendacion de oferta",
        "estrategia de oferta", "oferta sugerida", "strike", "walk-away",
    ])
    print(f"  8. Briefing incluye seccion Oferta Recomendada: {'✅' if c8 else '❌'}")

    print()
    score = sum([c1, c2, c3, c4, c5, c6, c7, c8])
    print(f"SCORE: {score}/8")

    # ===========================================================
    # Excerpts para revisión humana
    # ===========================================================
    print()
    print("=" * 70)
    print("EXCERPTS — ValueEstimator (primeras 800 chars)")
    print("=" * 70)
    print(ve_out[:800])

    print()
    print("=" * 70)
    print("EXCERPTS — OfferStrategist (primeras 1000 chars)")
    print("=" * 70)
    print(os_out[:1000])

    print()
    print("=" * 70)
    print("EXCERPTS — ContextualGlossaryAgent (primeras 1200 chars)")
    print("=" * 70)
    print(briefing_out[:1200])

    return 0 if score >= 7 else 1


if __name__ == "__main__":
    sys.exit(main())