From 7e72d073293124afab19174deec2aade8c350e73 Mon Sep 17 00:00:00 2001
From: Vadym Samoilenko <vadymsamoilenko@oliver.agency>
Date: Mon, 23 Mar 2026 19:17:36 +0000
Subject: [PATCH] Fix AI loop hanging: add asyncio.wait_for timeouts on LLM
 calls

The autonomous conversation loop could hang indefinitely because
self.response_timeout=30 was defined but never used in wait_for().

- autonomous_conversation_controller: wrap generate_persona_response()
  with asyncio.wait_for(timeout=120s); 30s was too short for production
  LLMs, raised to 120s; TimeoutError returns an error dict so the loop
  can continue or count toward consecutive_silence limit
- conversation_decision_service: add asyncio.wait_for(timeout=60s)
  around LLMService.generate_content() for the decision call; add
  asyncio import and explicit TimeoutError handling

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 .../autonomous_conversation_controller.py     | 29 ++++++++++++-------
 .../services/conversation_decision_service.py | 17 +++++++----
 2 files changed, 30 insertions(+), 16 deletions(-)

diff --git a/backend/app/services/autonomous_conversation_controller.py b/backend/app/services/autonomous_conversation_controller.py
index 8bab940b..5c66aa53 100755
--- a/backend/app/services/autonomous_conversation_controller.py
+++ b/backend/app/services/autonomous_conversation_controller.py
@@ -32,7 +32,7 @@ class AutonomousConversationController:
         # Timing configuration
         self.min_delay_between_actions = 3  # seconds
         self.max_delay_between_actions = 10  # seconds
-        self.response_timeout = 30  # seconds
+        self.response_timeout = 120  # seconds (LLM calls can be slow in production)
         
         # Edge case tracking
         self.consecutive_silence_count = 0
@@ -670,18 +670,25 @@ class AutonomousConversationController:
             messages = await FocusGroup.get_messages(self.focus_group_id)
             recent_messages = messages[-20:] if len(messages) > 20 else messages
             
-            # Generate response
+            # Generate response with timeout to prevent infinite hang
             try:
-                response_text = await generate_persona_response(
-                    persona=persona,
-                    current_topic=topic,
-                    previous_messages=recent_messages,
-                    temperature=0.7,
-                    focus_group_id=self.focus_group_id,
-                    llm_model=llm_model,
-                    reasoning_effort=reasoning_effort,
-                    verbosity=verbosity
+                response_text = await asyncio.wait_for(
+                    generate_persona_response(
+                        persona=persona,
+                        current_topic=topic,
+                        previous_messages=recent_messages,
+                        temperature=0.7,
+                        focus_group_id=self.focus_group_id,
+                        llm_model=llm_model,
+                        reasoning_effort=reasoning_effort,
+                        verbosity=verbosity
+                    ),
+                    timeout=self.response_timeout
                 )
+            except asyncio.TimeoutError:
+                error_msg = f"Participant response generation timed out after {self.response_timeout}s"
+                self.logger.error(f"⏱️ {error_msg}")
+                return {"error": error_msg}
             except Exception as e:
                 self.logger.error(f"Error in generate_persona_response: {str(e)}")
                 import traceback
diff --git a/backend/app/services/conversation_decision_service.py b/backend/app/services/conversation_decision_service.py
index 11e9983e..3fc3daac 100755
--- a/backend/app/services/conversation_decision_service.py
+++ b/backend/app/services/conversation_decision_service.py
@@ -4,6 +4,7 @@ Uses LLM to make intelligent decisions about conversation flow, participant sele
 """
 
 from typing import Dict, Any, Optional, List
+import asyncio
 import json
 from app.services.llm_service import LLMService, LLMServiceError
 from app.services.conversation_context_service import ConversationContextService
@@ -58,12 +59,15 @@ class ConversationDecisionService:
             focus_group = await FocusGroup.find_by_id(focus_group_id)
             llm_model = focus_group.get('llm_model') if focus_group else None
             
-            # Get LLM decision
+            # Get LLM decision with timeout to prevent infinite hang
             try:
-                response = await LLMService.generate_content(
-                    prompt=prompt,
-                    temperature=temperature,
-                    model_name=llm_model
+                response = await asyncio.wait_for(
+                    LLMService.generate_content(
+                        prompt=prompt,
+                        temperature=temperature,
+                        model_name=llm_model
+                    ),
+                    timeout=60
                 )
                 
                 # Parse the JSON response
@@ -92,6 +96,9 @@ class ConversationDecisionService:
                 
                 return decision
                 
+            except asyncio.TimeoutError:
+                print(f"⏱️ LLM decision timed out after 60s")
+                raise ConversationDecisionError("LLM decision timed out after 60s")
             except LLMServiceError as e:
                 print(f"❌ LLM Service Error: {str(e)}")
                 raise ConversationDecisionError(f"Error getting LLM decision: {str(e)}")