Spaces:
Runtime error
Runtime error
File size: 13,209 Bytes
b8ee1a5 49d8cb8 b8ee1a5 49d8cb8 b8ee1a5 49d8cb8 b8ee1a5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 |
# libraries imports
import yaml
import asyncio
from datetime import datetime
# from phoenix.otel import register
from llama_index.core.workflow import Context
from src.llms.gemini_2_flash import create_gemini
from llama_index.core.tools import QueryEngineTool
from llama_index.core.prompts import PromptTemplate
from llama_index.core.callbacks import CallbackManager
from llama_index.utils.workflow import draw_all_possible_flows
from llama_index.core.callbacks import CallbackManager, LlamaDebugHandler
from openinference.instrumentation.llama_index import LlamaIndexInstrumentor
from llama_index.core.agent.workflow import AgentWorkflow, ReActAgent, FunctionAgent
from llama_index.core.agent.workflow import AgentWorkflow, ToolCallResult, AgentStream
# Custom imports
from src.tools.web_search import search_tool
from src.tools.visit_webpage import visit_webpage_tool
from src.tools.query_on_url import Get_info_from_url_tool
from dotenv import load_dotenv
import os
# Load environment variables
load_dotenv()
class CustomDebugHandler(LlamaDebugHandler):
"""Custom debug handler for better traceability"""
def __init__(self):
super().__init__()
self.session_id = datetime.now().strftime("%Y%m%d_%H%M%S")
def on_event_start(self, event_type, payload=None, event_id="", **kwargs):
timestamp = datetime.now().strftime("%H:%M:%S")
print(f"\nπ [{timestamp}] Event Started: {event_type}")
if payload:
print(f" π Payload: {payload}")
super().on_event_start(event_type, payload, event_id, **kwargs)
def on_event_end(self, event_type, payload=None, event_id="", **kwargs):
timestamp = datetime.now().strftime("%H:%M:%S")
print(f"\nβ
[{timestamp}] Event Completed: {event_type}")
super().on_event_end(event_type, payload, event_id, **kwargs)
def create_callback_manager():
"""Create a callback manager with custom handlers"""
debug_handler = CustomDebugHandler()
callback_manager = CallbackManager([debug_handler])
return callback_manager
def get_agent_name_enhanced(ev, workflow):
"""Enhanced agent name detection with multiple fallback strategies"""
# Strategy 1: Direct attribute check
for attr in ['agent_name', 'name', 'sender', 'agent']:
if hasattr(ev, attr):
value = getattr(ev, attr)
if value and isinstance(value, str):
return value
# Strategy 2: Check source object
if hasattr(ev, 'source'):
source = ev.source
for attr in ['agent_name', 'name', 'id']:
if hasattr(source, attr):
value = getattr(source, attr)
if value and isinstance(value, str):
return value
# Strategy 3: Check workflow context or metadata
if hasattr(ev, 'metadata') and ev.metadata:
if 'agent_name' in ev.metadata:
return ev.metadata['agent_name']
# Strategy 4: Try to infer from workflow state
if hasattr(workflow, '_current_agent') and workflow._current_agent:
return workflow._current_agent
# Strategy 5: Check event type patterns
event_type = type(ev).__name__
if 'Agent' in event_type:
return f"Agent_{event_type}"
return "UnknownAgent"
def format_output_message(agent_name, message_type, content, timestamp=None):
"""Format output messages consistently"""
if timestamp is None:
timestamp = datetime.now().strftime("%H:%M:%S")
separator = "=" * 60
header = f"[{timestamp}] {agent_name} - {message_type}"
return f"\n{separator}\n{header}\n{separator}\n{content}\n{separator}\n"
def create_custom_react_prompt(agent_config):
"""Create a properly formatted ReAct prompt that maintains the required structure"""
# Extract the core mission and instructions from your config
core_instructions = agent_config["system_prompt"]
# Create a ReAct-compatible prompt that integrates your instructions
react_prompt = f"""You are a helpful AI assistant that can use tools to answer questions.
{core_instructions}
When responding, you must follow this exact format:
Thought: I need to think about what the user is asking and determine if I need to use any tools.
Action: [tool_name if using a tool, or skip this line if not using a tool]
Action Input: [tool input in JSON format if using a tool, or skip this line if not using a tool]
Observation: [This will be filled in by the tool response]
Continue this Thought/Action/Action Input/Observation cycle until you have enough information to provide a final answer.
When you have enough information, provide your final response in this format:
Thought: I now have enough information to provide a complete answer.
Answer: [Your final answer here]
Remember to always start with a Thought and follow the exact format above."""
return react_prompt
async def main():
# Create callback manager
callback_manager = create_callback_manager()
# phoenix handler
# Register the tracer provider (connects to OpenTelemetry)
# tracer_provider = register()
# Instrument LlamaIndex with OpenInference
# LlamaIndexInstrumentor().instrument(tracer_provider=tracer_provider)
# Create LLM with callback manager
llm = create_gemini()
if hasattr(llm, 'callback_manager'):
llm.callback_manager = callback_manager
def load_config(file_path):
"""Load configuration from a YAML file."""
try:
with open(file_path, 'r') as file:
return yaml.safe_load(file)
except Exception as e:
print(f"Error loading config file: {e}")
return {}
# Load configuration
config = load_config('src/agents/prompts.yaml')
# Create agents with proper prompt integration
try:
# Manager Agent
manager_agent = ReActAgent(
name=config["manager_agent"]["name"],
description=config["manager_agent"]["description"],
tools=[],
llm=llm,
callback_manager=callback_manager,
verbose=True, # Enable verbose mode for debugging
can_handoff_to=["product_hunter_agent", "product_investigator_agent", "trivial_search_agent"],
)
# Create custom prompt template for manager
manager_prompt = create_custom_react_prompt(config["manager_agent"])
manager_agent.update_prompts({
"react_header": PromptTemplate(manager_prompt)
})
# Product Hunter Agent
product_hunter_agent = ReActAgent(
name=config["product_hunter_agent"]["name"],
description=config["product_hunter_agent"]["description"],
tools=[search_tool, visit_webpage_tool, Get_info_from_url_tool],
llm=llm,
callback_manager=callback_manager,
verbose=True,
can_handoff_to=["manager_agent"],
# can_be_handed_off_by=["manager_agent"]
)
hunter_prompt = create_custom_react_prompt(config["product_hunter_agent"])
product_hunter_agent.update_prompts({
"react_header": PromptTemplate(hunter_prompt)
})
# Trivial Search Agent
trivial_search_agent = ReActAgent(
name=config["trivial_search_agent"]["name"],
description=config["trivial_search_agent"]["description"],
tools=[search_tool, visit_webpage_tool, Get_info_from_url_tool],
llm=llm,
callback_manager=callback_manager,
verbose=True,
can_handoff_to=["manager_agent"],
)
trivial_prompt = create_custom_react_prompt(config["trivial_search_agent"])
trivial_search_agent.update_prompts({
"react_header": PromptTemplate(trivial_prompt)
})
# Product Investigator Agent
product_investigator_agent = ReActAgent(
name=config["product_investigator_agent"]["name"],
description=config["product_investigator_agent"]["description"],
tools=[search_tool, visit_webpage_tool, Get_info_from_url_tool],
llm=llm,
callback_manager=callback_manager,
verbose=True,
can_handoff_to=["manager_agent"],
)
investigator_prompt = create_custom_react_prompt(config["product_investigator_agent"])
product_investigator_agent.update_prompts({
"react_header": PromptTemplate(investigator_prompt)
})
print("β
All agents created successfully!")
except Exception as e:
print(f"β Error creating agents: {e}")
import traceback
traceback.print_exc()
return
# Create workflow with callback manager
try:
workflow = AgentWorkflow(
agents=[manager_agent, product_hunter_agent,
product_investigator_agent, trivial_search_agent],
root_agent="manager_agent",
# callback_manager=callback_manager,
verbose=True,
)
print("β
Workflow created successfully!")
except Exception as e:
print(f"β Error creating workflow: {e}")
import traceback
traceback.print_exc()
return
# To keep memory
ctx = Context(workflow)
# Test prompts
prompt = "I want to buy a wired headset for gaming for 500 USD or less."
print(f"\nπ― Starting Shopping Assistant Workflow")
print(f"π Query: {prompt}")
print(f"β° Session Started: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
try:
handler = workflow.run(
user_msg=prompt,
ctx=ctx
)
# Create output file with timestamp
output_filename = f"agent_output_{datetime.now().strftime('%Y%m%d_%H%M%S')}.txt"
# Agent name mapping for better identification
agent_mapping = {
agent.name: agent.name for agent in [
manager_agent, product_hunter_agent, trivial_search_agent,
product_investigator_agent,
]
}
print(f"\nπ Logging to: {output_filename}")
with open(output_filename, "w", encoding="utf-8") as f:
# Write session header
session_header = f"""
Shopping Assistant Session Log
==============================
Session ID: {datetime.now().strftime('%Y%m%d_%H%M%S')}
Start Time: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
Query: {prompt}
Active Agents: {', '.join(agent_mapping.keys())}
==============================
"""
f.write(session_header)
print(session_header)
current_agent = "Unknown"
async for ev in handler.stream_events():
timestamp = datetime.now().strftime("%H:%M:%S")
agent_name = get_agent_name_enhanced(ev, workflow)
# Update current agent tracking
if agent_name != "UnknownAgent":
current_agent = agent_name
else:
agent_name = current_agent
if isinstance(ev, ToolCallResult):
tool_message = format_output_message(
agent_name,
"TOOL EXECUTION",
f"π Tool: {ev.tool_name}\n"
f"π₯ Input: {ev.tool_kwargs}\n"
f"π€ Output: {str(ev.tool_output)[:500]}{'...' if len(str(ev.tool_output)) > 500 else ''}",
timestamp
)
print(tool_message)
f.write(tool_message)
elif isinstance(ev, AgentStream):
delta = getattr(ev, "delta", "")
if delta.strip(): # Only log non-empty deltas
stream_message = f"[{timestamp}] {agent_name} π: {delta}"
print(stream_message, end="", flush=True)
f.write(stream_message)
# Write session footer
session_footer = f"""
==============================
Session Completed: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
==============================
"""
f.write(session_footer)
print(session_footer)
print("\nπ Workflow execution completed!")
# Get final response
try:
resp = await handler
final_response = format_output_message(
"FINAL_RESPONSE",
"WORKFLOW RESULT",
str(resp)
)
print(final_response)
# Append final response to file
with open(output_filename, "a", encoding="utf-8") as f:
f.write(final_response)
except Exception as e:
error_message = f"β Error getting final response: {str(e)}"
print(error_message)
with open(output_filename, "a", encoding="utf-8") as f:
f.write(f"\n{error_message}\n")
except Exception as e:
print(f"β Error running workflow: {e}")
import traceback
traceback.print_exc()
if __name__ == "__main__":
asyncio.run(main()) |