feat(bot): 5-phase strategy upgrade — edge neto, families, GNews priority, regimes

Phase 1 — Edge neto real (paper.py, bayesian.py, risk/manager.py, db.py):
- Trade records now store edge_gross, edge_net, prior_prob, final_prob,
  mid_price, spread_estimate, commission, family_key
- edge_net = edge_gross - SPREAD_ESTIMATE(0.02) - COMMISSION_RATE(0.02)
  NOTE: both constants are heuristics, not exact Polymarket exchange costs
- Execution gate changed from edge_gross > MIN_EDGE to edge_net > regime_min_edge

Phase 2 — Market families (polymarket.py):
- market_family_key(market) groups related markets:
    texas-republican-2026, fed-april-2026, openai-2026, etc.
- At most 1 trade per family per cycle; occupied_families propagated via main.py
- Family key logged on every TRADE and SKIP line

Phase 3 — GNews priority (news.py, bayesian.py, main.py):
- NewsClient.get_freshness() returns 1.0/0.75/0.40/0.10 by cache age
- gnews_priority(market, news) = uncertainty × volume_score × freshness
- Politics markets sorted by priority DESC before eval so best markets get
  the 5-query/cycle GNews budget first

Phase 4 — Regime min-edge by category/horizon (bayesian.py):
- politics >60d → 0.12, 30-60d → 0.10, <30d → 0.08
- tech / crypto/finance → 0.10
- All thresholds applied to edge_net (not edge_gross)

Phase 5 — Observability (bayesian.py, main.py):
- Structured skip labels: SKIP_UNSUPPORTED, SKIP_NO_SIGNALS,
  SKIP_PRIOR_EXTREME, SKIP_FAMILY, SKIP_GNEWS_PRIORITY, SKIP_EDGE_NET
- TRADE lines now include family_key, edge_gross, edge_net, regime_min, days
- schema.sql: 8 new cols on trades, 7 new cols on signals (via ALTER TABLE IF NOT EXISTS)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

bot/data/polymarket.py

@@ -5,6 +5,7 @@ Docs: https://docs.polymarket.com
 import asyncio
 import logging
 import os
+import re
 from dataclasses import dataclass, field
 from datetime import datetime, timezone, timedelta
 from typing import Optional
@@ -15,6 +16,158 @@ log = logging.getLogger(__name__)
 POLYMARKET_API = "https://clob.polymarket.com"
 GAMMA_API = "https://gamma-api.polymarket.com"
 
+# ─────────────────────────────────────────────────────────────────────────────
+# Phase 2 — Market family classification helpers
+# Used by market_family_key() below.
+# ─────────────────────────────────────────────────────────────────────────────
+
+_YEAR_RE = re.compile(r"\b(202\d|203\d)\b")
+_MONTH_RE = re.compile(
+    r"\b(january|february|march|april|may|june|july|august|"
+    r"september|october|november|december)\b",
+    re.IGNORECASE,
+)
+_FED_TRIGGER_RE = re.compile(
+    r"\b(federal reserve|interest rate|bps|basis point|fed\s+(rate|meeting|decision))",
+    re.IGNORECASE,
+)
+_US_STATE_RE = re.compile(
+    r"\b(Alabama|Alaska|Arizona|Arkansas|California|Colorado|Connecticut|"
+    r"Delaware|Florida|Georgia|Hawaii|Idaho|Illinois|Indiana|Iowa|Kansas|"
+    r"Kentucky|Louisiana|Maine|Maryland|Massachusetts|Michigan|Minnesota|"
+    r"Mississippi|Missouri|Montana|Nebraska|Nevada|New\s+Hampshire|"
+    r"New\s+Jersey|New\s+Mexico|New\s+York|North\s+Carolina|North\s+Dakota|"
+    r"Ohio|Oklahoma|Oregon|Pennsylvania|Rhode\s+Island|South\s+Carolina|"
+    r"South\s+Dakota|Tennessee|Texas|Utah|Vermont|Virginia|Washington|"
+    r"West\s+Virginia|Wisconsin|Wyoming)\b",
+    re.IGNORECASE,
+)
+_PARTY_RE = re.compile(r"\b(Republican|Democrats?|Democratic|GOP)\b", re.IGNORECASE)
+_ELECTION_TYPE_RE = re.compile(
+    r"\b(presidential|president|mayoral|mayor|gubernatorial|governor|"
+    r"senate|congress(?:ional)?|primary|election)\b",
+    re.IGNORECASE,
+)
+
+# Ordered list of (pattern, place_slug) for named non-US locations.
+# Checked after US-state patterns so US city/state names don't shadow these.
+_NAMED_PLACES: list[tuple[re.Pattern, str]] = [
+    (re.compile(r"\bColomb", re.IGNORECASE),            "colombia"),
+    (re.compile(r"\bSeoul\b", re.IGNORECASE),           "seoul"),
+    (re.compile(r"\bBusan\b", re.IGNORECASE),           "busan"),
+    (re.compile(r"\bGyeonggi\b", re.IGNORECASE),        "gyeonggi"),
+    (re.compile(r"\bChungcheong", re.IGNORECASE),       "chungcheong"),
+    (re.compile(r"\bSouth\s+Korean?\b", re.IGNORECASE), "south-korea"),
+    (re.compile(r"\bLos\s+Angeles\b", re.IGNORECASE),   "los-angeles"),
+    (re.compile(r"\bCuba\b", re.IGNORECASE),            "cuba"),
+    (re.compile(r"\bLebanon\b", re.IGNORECASE),         "lebanon"),
+    (re.compile(r"\bIsrael\b", re.IGNORECASE),          "israel"),
+    (re.compile(r"\bUkraine\b", re.IGNORECASE),         "ukraine"),
+    (re.compile(r"\bRussia\b", re.IGNORECASE),          "russia"),
+]
+
+# Ordered list of (pattern, company_slug) for tech/company markets.
+_NAMED_COMPANIES: list[tuple[re.Pattern, str]] = [
+    (re.compile(r"\bopenai\b", re.IGNORECASE),     "openai"),
+    (re.compile(r"\banthropic\b", re.IGNORECASE),  "anthropic"),
+    (re.compile(r"\bnvidia\b", re.IGNORECASE),     "nvidia"),
+    (re.compile(r"\bapple\b", re.IGNORECASE),      "apple"),
+    (re.compile(r"\bmicrosoft\b", re.IGNORECASE),  "microsoft"),
+    (re.compile(r"\bgoogle\b", re.IGNORECASE),     "google"),
+    (re.compile(r"\btesla\b", re.IGNORECASE),      "tesla"),
+    # \bmeta\b does NOT match MetaMask (no word boundary mid-compound-word)
+    (re.compile(r"\bmeta\b", re.IGNORECASE),       "meta"),
+]
+
+
+def _end_month(market: "Market") -> str:
+    """Return market end_date formatted as YYYY-MM, or '' if unparseable."""
+    raw = market.end_date
+    if not raw:
+        return ""
+    try:
+        dt = datetime.fromisoformat(raw.replace("Z", "+00:00"))
+        return dt.strftime("%Y-%m")
+    except (ValueError, TypeError):
+        return ""
+
+
+def market_family_key(market: "Market") -> str:
+    """
+    Return a stable slug that groups related markets together.
+
+    Markets in the same family share an underlying event (same election,
+    same Fed meeting decision, same company).  The bot allows at most one
+    open position per family per cycle to avoid correlated exposure.
+
+    Priority order (first match wins):
+      1. Fed / interest-rate decision  →  fed-{month}-{year}
+      2. US state + party election      →  {state}-{party}-{year}
+      3. Named non-US city/country      →  {place}-{event_type}-{year}
+      4. Named tech company             →  {company}-{year}
+      5. Fallback                       →  {category}-{end_YYYY-MM}
+
+    Examples:
+      "Will Ken Paxton win the 2026 Texas Republican Primary"
+          → texas-republican-2026
+      "Will the Fed decrease rates by 25 bps after April 2026 meeting"
+          → fed-april-2026
+      "Will OpenAI IPO by December 31 2026?"
+          → openai-2026
+    """
+    q = market.question
+
+    # Prefer year from question text; fall back to end_date year if absent
+    year_m = _YEAR_RE.search(q)
+    if year_m:
+        year = year_m.group(1)
+    else:
+        end_m = _end_month(market)      # e.g. "2026-06"
+        year = end_m[:4] if end_m else "unknown"
+
+    # 1. Fed / interest-rate meeting
+    if _FED_TRIGGER_RE.search(q):
+        month_m = _MONTH_RE.search(q)
+        if month_m:
+            return f"fed-{month_m.group(1).lower()}-{year}"
+        return f"fed-{year}"
+
+    # 2. US state + party (primary, senate, governor, etc.)
+    state_m = _US_STATE_RE.search(q)
+    party_m = _PARTY_RE.search(q)
+    if state_m and party_m:
+        state = re.sub(r"\s+", "-", state_m.group(1).lower())
+        raw_party = party_m.group(1).lower()
+        # "democrat" prefix covers "democrat", "democrats", "democratic"
+        party = "democrat" if "democrat" in raw_party else "republican"
+        return f"{state}-{party}-{year}"
+
+    # 3. Named non-US city / country
+    for place_re, place_slug in _NAMED_PLACES:
+        if place_re.search(q):
+            etype_m = _ELECTION_TYPE_RE.search(q)
+            if etype_m:
+                raw_etype = etype_m.group(1).lower()
+                # Normalise synonyms
+                etype = {
+                    "president": "presidential",
+                    "mayor":     "mayoral",
+                    "governor":  "gubernatorial",
+                }.get(raw_etype, raw_etype)
+            else:
+                etype = "event"
+            return f"{place_slug}-{etype}-{year}"
+
+    # 4. Named tech company
+    for company_re, company_slug in _NAMED_COMPANIES:
+        if company_re.search(q):
+            return f"{company_slug}-{year}"
+
+    # 5. Fallback: category + end_date month
+    end_month = _end_month(market)
+    base = market.category if market.category else "misc"
+    return f"{base}-{end_month}" if end_month else f"{base}-{year}"
+
 
 @dataclass
 class Market: