Spaces:
Sleeping
Sleeping
File size: 14,377 Bytes
8f7a08e | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 | """
Tests for multi-agent planning and execution system.
"""
import pytest
import os
import json
from app.env import CRMQueryEnv
from app.tasks import get_task_by_id, get_all_task_ids
from app.multi_agent import (
PlannerAgent, ExecutorAgent, Coordinator,
Plan, PlanStep
)
class TestPlannerAgentInitialization:
"""Test planner agent initialization."""
def test_planner_initialization(self):
"""Test planner agent can be initialized."""
planner = PlannerAgent()
assert planner.model == "gpt-3.5-turbo"
assert planner.temperature == 0.0
def test_planner_with_custom_api_key(self):
"""Test planner with custom API key."""
api_key = "test-key"
planner = PlannerAgent(api_key=api_key)
assert planner.api_key == api_key
def test_planner_uses_deterministic_temperature(self):
"""Test planner uses temperature=0 for determinism."""
planner = PlannerAgent()
assert planner.temperature == 0.0
class TestPlannerFallback:
"""Test planner fallback behavior."""
def test_fallback_plan_generation(self):
"""Test that fallback plan is generated correctly."""
planner = PlannerAgent()
plan = planner._generate_fallback_plan(
task_id="test_task",
task_description="Test task description"
)
assert plan.task_id == "test_task"
assert plan.description == "Test task description"
assert len(plan.steps) > 0
assert isinstance(plan.steps[0], PlanStep)
def test_fallback_plan_structure(self):
"""Test fallback plan has correct structure."""
planner = PlannerAgent()
plan = planner._generate_fallback_plan("task_1", "Description")
assert plan.total_steps > 0
assert all(step.tool in ["search_customers", "search_orders", "search_tickets", "submit_answer"]
for step in plan.steps)
assert all(hasattr(step, 'rationale') for step in plan.steps)
assert all(hasattr(step, 'expected_output') for step in plan.steps)
class TestExecutorAgentInitialization:
"""Test executor agent initialization."""
def test_executor_initialization(self):
"""Test executor agent can be initialized."""
executor = ExecutorAgent()
assert executor.memory_hits == 0
assert executor.memory_misses == 0
assert isinstance(executor.execution_history, list)
assert len(executor.execution_history) == 0
def test_executor_has_memory_tracking(self):
"""Test executor has memory tracking fields."""
executor = ExecutorAgent()
assert hasattr(executor, 'retrieved_entities')
assert hasattr(executor, 'memory_hits')
assert hasattr(executor, 'memory_misses')
def test_executor_reset(self):
"""Test executor reset method."""
executor = ExecutorAgent()
executor.memory_hits = 5
executor.memory_misses = 3
executor.reset()
assert executor.memory_hits == 0
assert executor.memory_misses == 0
assert len(executor.execution_history) == 0
class TestPlanExecution:
"""Test plan execution against environment."""
def test_execute_simple_plan(self):
"""Test executing a simple plan."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
plan = Plan(
task_id=env.current_task_id,
description="Test plan",
total_steps=2,
reasoning="Test reasoning",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Find gold tier customers",
expected_output="List of customers"
),
PlanStep(
step_number=2,
tool="submit_answer",
arguments={"customer_ids": []},
rationale="Submit answer",
expected_output="Task completed"
)
]
)
result = executor.execute_plan(plan, env)
assert result["plan_id"] == env.current_task_id
assert result["steps_executed"] > 0
assert env.done
def test_execution_tracks_memory(self):
"""Test that execution tracks retrieved entities."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
plan = Plan(
task_id=env.current_task_id,
description="Memory tracking test",
total_steps=1,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Search",
expected_output="Results"
)
]
)
result = executor.execute_plan(plan, env)
assert len(executor.retrieved_entities['customers']) > 0
assert result["memory_efficiency"] > 0
def test_execution_stops_on_done(self):
"""Test execution stops when environment marks done."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
plan = Plan(
task_id=env.current_task_id,
description="Stop early test",
total_steps=3,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="submit_answer",
arguments={"customer_ids": ["C001"]},
rationale="Early submit",
expected_output="Done"
),
PlanStep(
step_number=2,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Should not execute",
expected_output="Should not reach"
),
PlanStep(
step_number=3,
tool="submit_answer",
arguments={"customer_ids": ["C002"]},
rationale="Also should not execute",
expected_output="Should not reach"
)
]
)
result = executor.execute_plan(plan, env)
# Should stop after first submit_answer
assert result["steps_executed"] == 1
assert env.done
class TestCoordinatorInitialization:
"""Test coordinator initialization."""
def test_coordinator_initialization(self):
"""Test coordinator can be initialized."""
coordinator = Coordinator()
assert hasattr(coordinator, 'planner')
assert hasattr(coordinator, 'executor')
assert isinstance(coordinator.planner, PlannerAgent)
assert isinstance(coordinator.executor, ExecutorAgent)
def test_coordinator_with_api_key(self):
"""Test coordinator with custom API key."""
api_key = "test-key"
coordinator = Coordinator(api_key=api_key)
assert coordinator.planner.api_key == api_key
class TestCoordinatorPipeline:
"""Test full pipeline execution."""
def test_pipeline_with_fallback(self):
"""Test pipeline with fallback plan."""
env = CRMQueryEnv()
env.reset()
coordinator = Coordinator(api_key="invalid-key-for-fallback")
# Pipeline should use fallback
result = coordinator.run_pipeline(env, max_iterations=1)
assert result["task_id"] == env.current_task_id
assert result["iterations"] == 1
assert len(result["plans"]) > 0 or "plan_error" in result
def test_pipeline_executes_plans(self):
"""Test that pipeline executes generated plans."""
env = CRMQueryEnv()
env.reset()
coordinator = Coordinator(api_key="invalid-key")
result = coordinator.run_pipeline(env, max_iterations=1)
# Even with errors, should attempt execution
assert "task_id" in result
assert "steps_taken" in result
def test_pipeline_requires_reset(self):
"""Test that pipeline requires environment reset."""
env = CRMQueryEnv()
# Don't reset
coordinator = Coordinator()
with pytest.raises(ValueError):
coordinator.run_pipeline(env)
def test_pipeline_result_structure(self):
"""Test pipeline result has expected structure."""
env = CRMQueryEnv()
env.reset()
coordinator = Coordinator(api_key="invalid-key")
result = coordinator.run_pipeline(env, max_iterations=1)
assert "task_id" in result
assert "iterations" in result
assert "execution_results" in result
assert "episode_reward" in result
assert "steps_taken" in result
class TestPlanStructure:
"""Test plan data structure."""
def test_plan_creation(self):
"""Test creating a plan object."""
plan = Plan(
task_id="test_task",
description="Test description",
total_steps=2,
reasoning="Test reasoning",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Find gold customers",
expected_output="List of customers"
)
]
)
assert plan.task_id == "test_task"
assert plan.total_steps == 2
assert len(plan.steps) == 1
def test_plan_model_dump(self):
"""Test plan can be serialized."""
plan = Plan(
task_id="test",
description="Test",
total_steps=1,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={},
rationale="Test",
expected_output="Test"
)
]
)
dumped = plan.model_dump()
assert isinstance(dumped, dict)
assert "task_id" in dumped
assert "steps" in dumped
assert len(dumped["steps"]) == 1
def test_plan_step_validation(self):
"""Test plan step has required fields."""
step = PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Test",
expected_output="Test output"
)
assert step.step_number == 1
assert step.tool == "search_customers"
assert isinstance(step.arguments, dict)
class TestMultiAgentMemory:
"""Test multi-agent memory tracking."""
def test_executor_tracks_retrieved_entities(self):
"""Test executor tracks retrieved entities."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
plan = Plan(
task_id=env.current_task_id,
description="Memory test",
total_steps=1,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Get customers",
expected_output="Customers"
)
]
)
executor.execute_plan(plan, env)
assert len(executor.retrieved_entities['customers']) > 0
def test_executor_memory_efficiency_calculation(self):
"""Test executor calculates memory efficiency."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
# Plan with multiple same queries
plan = Plan(
task_id=env.current_task_id,
description="Efficiency test",
total_steps=2,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="search_customers",
arguments={"tier": "Gold"},
rationale="Query 1",
expected_output="Result 1"
),
PlanStep(
step_number=2,
tool="search_orders",
arguments={"status": "Completed"},
rationale="Query 2",
expected_output="Result 2"
)
]
)
result = executor.execute_plan(plan, env)
assert "memory_efficiency" in result
assert result["memory_efficiency"] >= 0.0
class TestMultiAgentErrorHandling:
"""Test error handling in multi-agent system."""
def test_executor_handles_invalid_action(self):
"""Test executor handles invalid actions gracefully."""
env = CRMQueryEnv()
env.reset()
executor = ExecutorAgent()
plan = Plan(
task_id=env.current_task_id,
description="Error test",
total_steps=1,
reasoning="Test",
steps=[
PlanStep(
step_number=1,
tool="invalid_tool",
arguments={},
rationale="Invalid",
expected_output="Error"
)
]
)
result = executor.execute_plan(plan, env)
# Should handle gracefully
assert isinstance(result, dict)
def test_plan_missing_api_key(self):
"""Test plan generation handles missing API key gracefully."""
planner = PlannerAgent(api_key=None)
# Should use fallback when API key missing
plan = planner._generate_fallback_plan("task", "description")
assert isinstance(plan, Plan)
assert len(plan.steps) > 0
|