fix(extractor): streaming with guardrails + catalyst_type normalization

- Switch Ollama calls from non-streaming to streaming with early termination
- Add loop detection, max token limit, and stall timeout guards
- Add catalyst_type alias normalizer to handle model hallucinations
- Add explicit enum values in extraction prompt for catalyst_type
- Add streaming config knobs to OllamaConfig

services/extractor/schemas.py

@@ -194,6 +194,39 @@ def validate_extraction(
 # Normalize model output before validation
 # ---------------------------------------------------------------------------
 
+_CATALYST_ALIASES: dict[str, str] = {
+    "strategic pivot": "other",
+    "strategic": "other",
+    "restructuring": "other",
+    "partnership": "other",
+    "acquisition": "m_and_a",
+    "merger": "m_and_a",
+    "buyout": "m_and_a",
+    "lawsuit": "legal",
+    "regulation": "legal",
+    "regulatory": "legal",
+    "upgrade": "rating_change",
+    "downgrade": "rating_change",
+    "price target": "rating_change",
+    "inflation": "macro",
+    "interest rate": "macro",
+    "interest rates": "macro",
+    "tariff": "macro",
+    "tariffs": "macro",
+    "launch": "product",
+    "product launch": "product",
+    "revenue": "earnings",
+    "profit": "earnings",
+    "guidance": "earnings",
+    "supply": "supply_chain",
+    "shortage": "supply_chain",
+}
+
+_VALID_CATALYSTS = frozenset({
+    "earnings", "product", "legal", "macro",
+    "supply_chain", "m_and_a", "rating_change", "other",
+})
+
 _HORIZON_MAP: dict[str, str] = {
     "long-term": "90d_plus",
     "long": "90d_plus",
@@ -233,6 +266,11 @@ def _normalize_extraction_data(data: dict[str, Any]) -> dict[str, Any]:
                 mapped = _HORIZON_MAP.get(horizon.lower().strip())
                 if mapped:
                     comp["impact_horizon"] = mapped
+            # Map catalyst_type alternatives
+            cat = comp.get("catalyst_type", "")
+            if isinstance(cat, str) and cat.lower().strip() not in _VALID_CATALYSTS:
+                mapped_cat = _CATALYST_ALIASES.get(cat.lower().strip(), "other")
+                comp["catalyst_type"] = mapped_cat
 
     return data