Compare commits

...

2 Commits

Author SHA1 Message Date
Devin AI
3e914133bd Fix lint issues: remove unused imports from test file
- Remove unused pytest import
- Remove unused patch import
- Remove unused MAX_FILE_NAME_LENGTH import
- All ruff checks now pass locally

Co-Authored-By: João <joao@crewai.com>
2025-07-29 23:03:55 +00:00
Devin AI
f0949f90dd Fix Windows path length issue in memory storage
- Replace agent.role with agent.id (UUID) in RAGStorage directory naming
- UUIDs are guaranteed short (36 chars) and filesystem-safe
- Remove unused _sanitize_role method since UUIDs don't need sanitization
- Add comprehensive tests for Windows path length scenarios
- Add test case for long agent roles in existing memory tests
- Maintains backward compatibility while fixing Windows 260-char limit

Fixes #3236

Co-Authored-By: João <joao@crewai.com>
2025-07-29 23:00:35 +00:00
3 changed files with 226 additions and 6 deletions

View File

@@ -44,7 +44,7 @@ class RAGStorage(BaseRAGStorage):
):
super().__init__(type, allow_reset, embedder_config, crew)
agents = crew.agents if crew else []
agents = [self._sanitize_role(agent.role) for agent in agents]
agents = [str(agent.id) for agent in agents]
agents = "_".join(agents)
self.agents = agents
self.storage_file_name = self._build_storage_file_name(type, agents)
@@ -74,11 +74,6 @@ class RAGStorage(BaseRAGStorage):
)
logging.info(f"Collection found or created: {self.collection}")
def _sanitize_role(self, role: str) -> str:
"""
Sanitizes agent roles to ensure valid directory names.
"""
return role.replace("\n", "").replace(" ", "_").replace("/", "_")
def _build_storage_file_name(self, type: str, file_name: str) -> str:
"""

View File

@@ -160,3 +160,41 @@ def test_save_and_search(short_term_memory):
find = short_term_memory.search("test value", score_threshold=0.01)[0]
assert find["context"] == memory.data, "Data value mismatch."
assert find["metadata"]["agent"] == "test_agent", "Agent value mismatch."
def test_memory_with_long_agent_role():
"""Test that memory works correctly with very long agent roles."""
very_long_role = (
"Senior Equity Research Analyst specializing in corporate fundamentals and industry dynamics "
"with expertise in financial modeling, valuation techniques, and market analysis for "
"technology, healthcare, and consumer discretionary sectors, responsible for generating "
"comprehensive investment recommendations and detailed research reports"
)
agent = Agent(
role=very_long_role,
goal="Search relevant data and provide results",
backstory="You are a researcher at a leading tech think tank.",
tools=[],
verbose=True,
)
task = Task(
description="Perform a search on specific topics.",
expected_output="A list of relevant URLs based on the search query.",
agent=agent,
)
memory = ShortTermMemory(crew=Crew(agents=[agent], tasks=[task]))
test_data = "Test memory data for long role agent"
test_metadata = {"task": "test_task"}
memory.save(
value=test_data,
metadata=test_metadata,
agent=very_long_role,
)
results = memory.search("Test memory", score_threshold=0.01)
assert isinstance(results, list), "Search should return a list even with long agent roles"

View File

@@ -0,0 +1,187 @@
from crewai.agent import Agent
from crewai.crew import Crew
from crewai.memory.storage.rag_storage import RAGStorage
from crewai.task import Task
def test_long_agent_role_memory_storage():
"""Test that agents with very long roles don't exceed Windows path limits."""
very_long_role = (
"Senior Equity Research Analyst specializing in corporate fundamentals and industry dynamics "
"with expertise in financial modeling, valuation techniques, and market analysis for "
"technology, healthcare, and consumer discretionary sectors, responsible for generating "
"comprehensive investment recommendations and detailed research reports"
)
agent = Agent(
role=very_long_role,
goal="Analyze market trends and provide investment insights",
backstory="You are an experienced financial analyst with deep market knowledge.",
verbose=True,
)
task = Task(
description="Analyze the current market conditions.",
expected_output="A comprehensive market analysis report.",
agent=agent,
)
crew = Crew(agents=[agent], tasks=[task])
rag_storage = RAGStorage(type="short_term", crew=crew)
assert len(rag_storage.storage_file_name) <= 260, (
f"Storage path too long: {len(rag_storage.storage_file_name)} characters"
)
assert str(agent.id) in rag_storage.agents, (
"Agent UUID should be used in storage path"
)
def test_multiple_agents_with_long_roles():
"""Test that multiple agents with long roles create valid storage paths."""
long_roles = [
"Senior Investment Advisor specializing in equity portfolio strategy and client tailored recommendations",
"Financial Communications Specialist skilled in distilling complex analysis into concise client-facing investment reports",
"Registered Investment Advisor (RIA) specializing in equity portfolio strategy and client-tailored recommendations"
]
agents = []
for i, role in enumerate(long_roles):
agent = Agent(
role=role,
goal=f"Goal for agent {i+1}",
backstory=f"Backstory for agent {i+1}",
verbose=True,
)
agents.append(agent)
tasks = [
Task(
description=f"Task {i+1}",
expected_output=f"Output {i+1}",
agent=agent,
)
for i, agent in enumerate(agents)
]
crew = Crew(agents=agents, tasks=tasks)
rag_storage = RAGStorage(type="short_term", crew=crew)
assert len(rag_storage.storage_file_name) <= 260, (
f"Storage path too long: {len(rag_storage.storage_file_name)} characters"
)
for agent in agents:
assert str(agent.id) in rag_storage.agents, (
f"Agent UUID {agent.id} should be in storage path"
)
def test_memory_functionality_with_long_roles():
"""Test that memory save/search functionality works with long agent roles."""
long_role = (
"Senior Equity Research Analyst specializing in corporate fundamentals and industry dynamics "
"with expertise in financial modeling, valuation techniques, and market analysis"
)
agent = Agent(
role=long_role,
goal="Analyze market trends",
backstory="You are an experienced analyst.",
verbose=True,
)
task = Task(
description="Analyze market conditions.",
expected_output="Market analysis report.",
agent=agent,
)
crew = Crew(agents=[agent], tasks=[task])
rag_storage = RAGStorage(type="short_term", crew=crew)
test_value = "Test memory content for long role agent"
test_metadata = {"task": "test_task", "agent": long_role}
rag_storage.save(value=test_value, metadata=test_metadata)
results = rag_storage.search(query="Test memory", limit=1, score_threshold=0.1)
assert isinstance(results, list), "Search should return a list"
def test_backward_compatibility_short_roles():
"""Test that short agent roles still work correctly."""
agent = Agent(
role="Researcher",
goal="Research topics",
backstory="You are a researcher.",
verbose=True,
)
task = Task(
description="Research a topic.",
expected_output="Research results.",
agent=agent,
)
crew = Crew(agents=[agent], tasks=[task])
rag_storage = RAGStorage(type="short_term", crew=crew)
assert str(agent.id) in rag_storage.agents, (
"Agent UUID should be used even for short roles"
)
assert len(rag_storage.storage_file_name) <= 260, (
f"Storage path should be valid: {len(rag_storage.storage_file_name)} characters"
)
def test_uuid_based_path_uniqueness():
"""Test that different agents with same role create different storage paths."""
role = "Data Analyst"
agent1 = Agent(
role=role,
goal="Analyze data",
backstory="You are an analyst.",
verbose=True,
)
agent2 = Agent(
role=role,
goal="Analyze data",
backstory="You are an analyst.",
verbose=True,
)
task1 = Task(
description="Analyze dataset 1.",
expected_output="Analysis 1.",
agent=agent1,
)
task2 = Task(
description="Analyze dataset 2.",
expected_output="Analysis 2.",
agent=agent2,
)
crew1 = Crew(agents=[agent1], tasks=[task1])
crew2 = Crew(agents=[agent2], tasks=[task2])
storage1 = RAGStorage(type="short_term", crew=crew1)
storage2 = RAGStorage(type="short_term", crew=crew2)
assert storage1.agents != storage2.agents, (
"Different agents should create different storage paths even with same role"
)
assert str(agent1.id) in storage1.agents
assert str(agent2.id) in storage2.agents
assert str(agent1.id) != str(agent2.id)