fix(agents): stop conflict alone from defaulting the verdict to Hold

- both manager prompts and both rating fields listed "materially conflicting"
  as a Hold condition, and every bull/bear and risk debate is conflicting, so
  a modest directional edge collapsed to Hold
- state one rule at all four sites: conflict alone is not a reason to Hold,
  commit to the stronger side sized by how decisively it wins, and Hold only
  when the evidence is still balanced after weighing or too thin for a call
- a modest bull edge now resolves to Overweight; a genuinely balanced debate
  still resolves to Hold #1321

tests/test_structured_agents.py

@@ -7,12 +7,14 @@ behavior we added for the Trader, Research Manager, and Sentiment Analyst
 so they share the same deterministic output shape.
 """
 
+import inspect
 from unittest.mock import MagicMock
 
 import pytest
 from pydantic import ValidationError
 
 from tradingagents.agents.analysts.sentiment_analyst import create_sentiment_analyst
+from tradingagents.agents.managers.portfolio_manager import create_portfolio_manager
 from tradingagents.agents.managers.research_manager import create_research_manager
 from tradingagents.agents.schemas import (
     PortfolioDecision,
@@ -484,3 +486,20 @@ class TestSentimentAnalystAgent:
         llm.with_structured_output.return_value = structured
         llm.invoke.return_value = MagicMock(content=plain)
         assert create_sentiment_analyst(llm)(_make_sentiment_state())["sentiment_report"] == plain
+
+
+@pytest.mark.unit
+@pytest.mark.parametrize("source", [
+    pytest.param(lambda: ResearchPlan.model_fields["recommendation"].description, id="ResearchPlan.recommendation"),
+    pytest.param(lambda: PortfolioDecision.model_fields["rating"].description, id="PortfolioDecision.rating"),
+    pytest.param(lambda: inspect.getsource(create_research_manager), id="research_manager prompt"),
+    pytest.param(lambda: inspect.getsource(create_portfolio_manager), id="portfolio_manager prompt"),
+])
+def test_conflict_alone_is_not_a_hold_trigger(source):
+    # The debate always contains conflicting arguments, so a Hold condition that
+    # conflict satisfies fires on every run and swallows directional calls
+    # (#1321). All four decision sites must state the same rule.
+    text = " ".join(source().split())
+    assert "conflict alone is not a reason to Hold" in text or \
+        "Conflicting arguments alone are not a reason to Hold" in text
+    assert "materially conflicting" not in text