File size: 15,600 Bytes
b30f068 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 | """
Supervisor agent for orchestrating task execution using LangGraph.
"""
import time
from typing import Dict, Any, List, Optional, Callable
from dataclasses import dataclass
from enum import Enum
try:
from langgraph.graph import StateGraph, END
from langgraph.checkpoint.memory import MemorySaver
LANGGRAPH_AVAILABLE = True
except ImportError:
LANGGRAPH_AVAILABLE = False
from .base_agent import BaseAgent, AgentResult, AgentState, AgentMessage
from src.utils.logging_config import logger
class TaskStatus(Enum):
"""Task execution status."""
PENDING = "pending"
IN_PROGRESS = "in_progress"
COMPLETED = "completed"
FAILED = "failed"
RETRYING = "retrying"
@dataclass
class Task:
"""Represents a task to be executed."""
id: str
name: str
agent_type: str
input_data: Dict[str, Any]
dependencies: List[str] = None
status: TaskStatus = TaskStatus.PENDING
result: Optional[AgentResult] = None
retry_count: int = 0
max_retries: int = 3
@dataclass
class WorkflowState:
"""State of the workflow execution."""
query: str
tasks: List[Task]
results: Dict[str, Any]
current_task: Optional[str] = None
error_message: Optional[str] = None
completed: bool = False
metadata: Dict[str, Any] = None
class SupervisorAgent(BaseAgent):
"""Supervisor agent that orchestrates task execution using LangGraph."""
def __init__(self, agent_id: Optional[str] = None, config: Optional[Dict[str, Any]] = None):
"""Initialize the supervisor agent."""
if not LANGGRAPH_AVAILABLE:
raise ImportError("LangGraph is required for SupervisorAgent. Install with: pip install langgraph")
super().__init__(agent_id, config)
# Worker agents registry
self.worker_agents: Dict[str, BaseAgent] = {}
# Workflow configuration
self.max_retries = self.config.get('max_retries', 3)
self.retry_delay = self.config.get('retry_delay', 1.0)
# Build the workflow graph
self._build_workflow_graph()
def _initialize(self) -> None:
"""Initialize supervisor-specific components."""
self.checkpointer = MemorySaver()
logger.info("Supervisor agent initialized with LangGraph")
def _build_workflow_graph(self) -> None:
"""Build the LangGraph workflow."""
# Create the state graph
workflow = StateGraph(WorkflowState)
# Add nodes for each step
workflow.add_node("parse_query", self._parse_query_node)
workflow.add_node("match_apis", self._match_apis_node)
workflow.add_node("execute_apis", self._execute_apis_node)
workflow.add_node("format_results", self._format_results_node)
workflow.add_node("evaluate_results", self._evaluate_results_node)
# Define the workflow edges
workflow.set_entry_point("parse_query")
workflow.add_edge("parse_query", "match_apis")
workflow.add_edge("match_apis", "execute_apis")
workflow.add_edge("execute_apis", "format_results")
workflow.add_edge("format_results", "evaluate_results")
workflow.add_edge("evaluate_results", END)
# Compile the graph
self.workflow = workflow.compile(checkpointer=self.checkpointer)
logger.info("LangGraph workflow compiled successfully")
def register_worker(self, agent_type: str, agent: BaseAgent) -> None:
"""
Register a worker agent.
Args:
agent_type: Type identifier for the agent
agent: Worker agent instance
"""
self.worker_agents[agent_type] = agent
logger.info(f"Registered worker agent: {agent_type} ({agent.agent_id})")
def execute(self, input_data: Dict[str, Any]) -> AgentResult:
"""
Execute the supervised workflow.
Args:
input_data: Input containing query and configuration
Returns:
AgentResult with workflow outcome
"""
start_time = time.time()
self.set_state(AgentState.RUNNING)
try:
# Validate input
if not self.validate_input(input_data):
raise ValueError("Invalid input data")
query = input_data.get('query', '')
if not query:
raise ValueError("Query is required")
# Initialize workflow state
initial_state = WorkflowState(
query=query,
tasks=[],
results={},
metadata=input_data.get('metadata', {})
)
# Execute the workflow
config = {"configurable": {"thread_id": f"workflow_{int(time.time())}"}}
final_state = self.workflow.invoke(initial_state, config)
execution_time = time.time() - start_time
if final_state.completed and not final_state.error_message:
self.set_state(AgentState.COMPLETED)
result = AgentResult(
agent_id=self.agent_id,
success=True,
data=final_state.results,
execution_time=execution_time,
metadata={
'tasks_completed': len([t for t in final_state.tasks if t.status == TaskStatus.COMPLETED]),
'workflow_state': final_state
}
)
else:
self.set_state(AgentState.FAILED)
result = AgentResult(
agent_id=self.agent_id,
success=False,
data=final_state.results,
error_message=final_state.error_message or "Workflow failed",
execution_time=execution_time,
metadata={'workflow_state': final_state}
)
self.log_execution(result)
return result
except Exception as e:
execution_time = time.time() - start_time
self.set_state(AgentState.FAILED)
result = AgentResult(
agent_id=self.agent_id,
success=False,
data=None,
error_message=str(e),
execution_time=execution_time
)
self.log_execution(result)
return result
def _parse_query_node(self, state: WorkflowState) -> WorkflowState:
"""Parse the input query."""
logger.info(f"Parsing query: {state.query}")
try:
# Execute query parsing using worker agent
parser_agent = self.worker_agents.get('query_parser')
if not parser_agent:
raise ValueError("Query parser agent not registered")
parse_result = parser_agent.execute({'query': state.query})
if parse_result.success:
state.results['parsed_query'] = parse_result.data
logger.info("Query parsing completed successfully")
else:
state.error_message = f"Query parsing failed: {parse_result.error_message}"
logger.error(state.error_message)
except Exception as e:
state.error_message = f"Query parsing error: {str(e)}"
logger.error(state.error_message)
return state
def _match_apis_node(self, state: WorkflowState) -> WorkflowState:
"""Match parsed query to available APIs."""
logger.info("Matching APIs")
if state.error_message:
return state
try:
# Get parsed query results
parsed_query = state.results.get('parsed_query')
if not parsed_query:
raise ValueError("No parsed query available")
# Execute API matching
matcher_agent = self.worker_agents.get('api_matcher')
if matcher_agent:
match_result = matcher_agent.execute({
'keywords': parsed_query.get('keywords', []),
'intent': parsed_query.get('intent')
})
if match_result.success:
state.results['api_matches'] = match_result.data
logger.info(f"Found {len(match_result.data.get('matches', []))} API matches")
else:
logger.warning(f"API matching failed: {match_result.error_message}")
state.results['api_matches'] = {'matches': []}
else:
logger.warning("API matcher agent not registered")
state.results['api_matches'] = {'matches': []}
except Exception as e:
state.error_message = f"API matching error: {str(e)}"
logger.error(state.error_message)
return state
def _execute_apis_node(self, state: WorkflowState) -> WorkflowState:
"""Execute matched API calls."""
logger.info("Executing API calls")
if state.error_message:
return state
try:
# Get API matches
api_matches = state.results.get('api_matches', {})
matches = api_matches.get('matches', [])
if not matches:
logger.info("No API matches to execute")
state.results['api_results'] = []
return state
# Execute API calls using executor agent
executor_agent = self.worker_agents.get('api_executor')
if not executor_agent:
raise ValueError("API executor agent not registered")
execution_result = executor_agent.execute({
'matches': matches,
'query_context': state.results.get('parsed_query')
})
if execution_result.success:
state.results['api_results'] = execution_result.data
logger.info(f"Executed {len(execution_result.data.get('results', []))} API calls")
else:
state.error_message = f"API execution failed: {execution_result.error_message}"
logger.error(state.error_message)
except Exception as e:
state.error_message = f"API execution error: {str(e)}"
logger.error(state.error_message)
return state
def _format_results_node(self, state: WorkflowState) -> WorkflowState:
"""Format the results for output."""
logger.info("Formatting results")
if state.error_message:
return state
try:
# Format results using formatter agent
formatter_agent = self.worker_agents.get('result_formatter')
if not formatter_agent:
# Basic formatting if no formatter agent
state.results['formatted_output'] = {
'query': state.query,
'results': state.results.get('api_results', []),
'metadata': state.metadata
}
logger.info("Applied basic result formatting")
return state
format_result = formatter_agent.execute({
'query': state.query,
'parsed_query': state.results.get('parsed_query'),
'api_matches': state.results.get('api_matches'),
'api_results': state.results.get('api_results'),
'metadata': state.metadata
})
if format_result.success:
state.results['formatted_output'] = format_result.data
logger.info("Result formatting completed successfully")
else:
state.error_message = f"Result formatting failed: {format_result.error_message}"
logger.error(state.error_message)
except Exception as e:
state.error_message = f"Result formatting error: {str(e)}"
logger.error(state.error_message)
return state
def _evaluate_results_node(self, state: WorkflowState) -> WorkflowState:
"""Evaluate the final results."""
logger.info("Evaluating results")
try:
# Basic evaluation - can be enhanced with evaluator agent
evaluator_agent = self.worker_agents.get('evaluator')
if evaluator_agent:
eval_result = evaluator_agent.execute({
'query': state.query,
'results': state.results,
'workflow_state': state
})
if eval_result.success:
state.results['evaluation'] = eval_result.data
# Check if results meet quality threshold
quality_score = eval_result.data.get('quality_score', 0.5)
if quality_score >= 0.7:
state.completed = True
logger.info(f"Workflow completed successfully (quality: {quality_score:.2f})")
else:
logger.warning(f"Results quality below threshold: {quality_score:.2f}")
state.completed = True # Complete anyway for now
else:
logger.warning(f"Result evaluation failed: {eval_result.error_message}")
state.completed = True # Complete anyway
else:
# Basic evaluation without evaluator agent
has_results = bool(state.results.get('api_results'))
state.completed = True
state.results['evaluation'] = {
'has_results': has_results,
'quality_score': 0.8 if has_results else 0.3,
'evaluation_method': 'basic'
}
logger.info(f"Basic evaluation completed: {'success' if has_results else 'limited results'}")
except Exception as e:
state.error_message = f"Result evaluation error: {str(e)}"
logger.error(state.error_message)
state.completed = True # Complete with error
return state
def get_capabilities(self) -> List[str]:
"""Get supervisor capabilities."""
return [
"workflow_orchestration",
"task_scheduling",
"agent_coordination",
"error_handling",
"result_aggregation"
]
def get_workflow_status(self) -> Dict[str, Any]:
"""Get current workflow status."""
return {
'registered_workers': list(self.worker_agents.keys()),
'workflow_available': LANGGRAPH_AVAILABLE,
'execution_history_count': len(self.execution_history)
}
|