github-mirrors/tradingagents
1
0
Fork 0
mirror of https://github.com/TauricResearch/TradingAgents.git synced 2026-09-23 21:18:26 +03:00
Code Issues Packages Projects Releases Wiki Activity

fix(agents): stop debate openers from rebutting a nonexistent argument

Browse Source 
- the first speaker in each debate round received an empty opponent response,
  yet the prompt demanded a rebuttal, so models fabricated the other side
- substitute an explicit opening marker when an opponent has not spoken, across
  all five debators (bull, bear, and the three risk analysts) #1176
This commit is contained in:
Yijia-Xiao
2026-08-30 06:18:59 +00:00
parent 9b98f09613
commit 539eae8fd6
7 changed files with 154 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
tradingagents/agents/risk_mgmt/aggressive_debator.py
View File

@@ -1,6 +1,7 @@
from tradingagents.agents.utils.agent_utils import (
get_instrument_context_from_state,
get_language_instruction,
opponent_argument_or_opening,
)
@@ -10,8 +11,12 @@ def create_aggressive_debator(llm):
history = risk_debate_state.get("history", "")
aggressive_history = risk_debate_state.get("aggressive_history", "")
current_conservative_response = risk_debate_state.get("current_conservative_response", "")
current_neutral_response = risk_debate_state.get("current_neutral_response", "")
current_conservative_response = opponent_argument_or_opening(
risk_debate_state.get("current_conservative_response", ""), "conservative analyst"
)
current_neutral_response = opponent_argument_or_opening(
risk_debate_state.get("current_neutral_response", ""), "neutral analyst"
)
market_research_report = state["market_report"]
sentiment_report = state["sentiment_report"]