Skip to content

πŸ™Œ OpenHands: AI-Driven Development

Set default condenser to `ConversationWindowCondenser` (#10031)

Calvin Smith39fff41dd4c0
diff --git a/openhands/cli/settings.py b/openhands/cli/settings.py
index de2b4c5f58..21cb1f5d89 100644
--- a/openhands/cli/settings.py
+++ b/openhands/cli/settings.py
@@ -23,7 +23,10 @@ from openhands.cli.utils import (
)
from openhands.controller.agent import Agent
from openhands.core.config import OpenHandsConfig
-from openhands.core.config.condenser_config import NoOpCondenserConfig
+from openhands.core.config.condenser_config import (
+ CondenserPipelineConfig,
+ ConversationWindowCondenserConfig,
+)
from openhands.core.config.utils import OH_DEFAULT_AGENT
from openhands.memory.condenser.impl.llm_summarizing_condenser import (
LLMSummarizingCondenserConfig,
@@ -472,12 +475,21 @@ async def modify_llm_settings_advanced(
agent_config = config.get_agent_config(config.default_agent)
if enable_memory_condensation:
- agent_config.condenser = LLMSummarizingCondenserConfig(
- llm_config=llm_config,
- type='llm',
+ agent_config.condenser = CondenserPipelineConfig(
+ type='pipeline',
+ condensers=[
+ ConversationWindowCondenserConfig(type='conversation_window'),
+ # Use LLMSummarizingCondenserConfig with the custom llm_config
+ LLMSummarizingCondenserConfig(
+ llm_config=llm_config, type='llm', keep_first=4, max_size=120
+ ),
+ ],
)
+
else:
- agent_config.condenser = NoOpCondenserConfig(type='noop')
+ agent_config.condenser = ConversationWindowCondenserConfig(
+ type='conversation_window'
+ )
config.set_agent_config(agent_config)
settings = await settings_store.load()
diff --git a/openhands/core/config/agent_config.py b/openhands/core/config/agent_config.py
index ef280ff965..53bee7b46b 100644
--- a/openhands/core/config/agent_config.py
+++ b/openhands/core/config/agent_config.py
@@ -2,7 +2,10 @@ from __future__ import annotations
from pydantic import BaseModel, ConfigDict, Field, ValidationError
-from openhands.core.config.condenser_config import CondenserConfig, NoOpCondenserConfig
+from openhands.core.config.condenser_config import (
+ CondenserConfig,
+ ConversationWindowCondenserConfig,
+)
from openhands.core.config.extended_config import ExtendedConfig
from openhands.core.logger import openhands_logger as logger
from openhands.utils.import_utils import get_impl
@@ -44,7 +47,11 @@ class AgentConfig(BaseModel):
enable_som_visual_browsing: bool = Field(default=True)
"""Whether to enable SoM (Set of Marks) visual browsing."""
condenser: CondenserConfig = Field(
- default_factory=lambda: NoOpCondenserConfig(type='noop')
+ # The default condenser is set to the conversation window condenser -- if
+ # we use NoOp and the conversation hits the LLM context length limit,
+ # the agent will generate a condensation request which will never be
+ # handled.
+ default_factory=lambda: ConversationWindowCondenserConfig()
)
extended: ExtendedConfig = Field(default_factory=lambda: ExtendedConfig({}))
"""Extended configuration for the agent."""
diff --git a/tests/unit/test_agents.py b/tests/unit/test_agents.py
index ffd2ab3498..32ad833612 100644
--- a/tests/unit/test_agents.py
+++ b/tests/unit/test_agents.py
@@ -249,6 +249,7 @@ def test_step_with_no_pending_actions(mock_state: State):
llm.completion = Mock(return_value=mock_response)
llm.is_function_calling_active = Mock(return_value=True) # Enable function calling
llm.is_caching_prompt_active = Mock(return_value=False)
+ llm.format_messages_for_llm = Mock(return_value=[]) # Mock message formatting
# Create agent with mocked LLM
config = AgentConfig()
@@ -269,6 +270,10 @@ def test_step_with_no_pending_actions(mock_state: State):
initial_user_message._source = EventSource.USER
mock_state.history = [initial_user_message]
+ # Mock the view returned by condenser
+ mock_view = View(events=mock_state.history)
+ mock_state.view = mock_view
+
action = agent.step(mock_state)
assert isinstance(action, MessageAction)
assert action.content == 'Task completed'
diff --git a/tests/unit/test_cli_settings.py b/tests/unit/test_cli_settings.py
index e0d7f60de5..758c8d70f8 100644
--- a/tests/unit/test_cli_settings.py
+++ b/tests/unit/test_cli_settings.py
@@ -19,16 +19,24 @@ from openhands.storage.settings.file_settings_store import FileSettingsStore
# Mock classes for condensers
class MockLLMSummarizingCondenserConfig:
- def __init__(self, llm_config, type):
+ def __init__(self, llm_config, type, keep_first=4, max_size=120):
self.llm_config = llm_config
self.type = type
+ self.keep_first = keep_first
+ self.max_size = max_size
-class MockNoOpCondenserConfig:
+class MockConversationWindowCondenserConfig:
def __init__(self, type):
self.type = type
+class MockCondenserPipelineConfig:
+ def __init__(self, type, condensers):
+ self.type = type
+ self.condensers = condensers
+
+
class TestDisplaySettings:
@pytest.fixture
def app_config(self):
@@ -467,7 +475,13 @@ class TestModifyLLMSettingsAdvanced:
'openhands.cli.settings.LLMSummarizingCondenserConfig',
MockLLMSummarizingCondenserConfig,
)
- @patch('openhands.cli.settings.NoOpCondenserConfig', MockNoOpCondenserConfig)
+ @patch(
+ 'openhands.cli.settings.ConversationWindowCondenserConfig',
+ MockConversationWindowCondenserConfig,
+ )
+ @patch(
+ 'openhands.cli.settings.CondenserPipelineConfig', MockCondenserPipelineConfig
+ )
async def test_modify_llm_settings_advanced_success(
self, mock_confirm, mock_session, mock_list_agents, app_config, settings_store
):
@@ -521,7 +535,10 @@ class TestModifyLLMSettingsAdvanced:
'openhands.cli.settings.LLMSummarizingCondenserConfig',
MockLLMSummarizingCondenserConfig,
)
- @patch('openhands.cli.settings.NoOpCondenserConfig', MockNoOpCondenserConfig)
+ @patch(
+ 'openhands.cli.settings.ConversationWindowCondenserConfig',
+ MockConversationWindowCondenserConfig,
+ )
async def test_modify_llm_settings_advanced_user_cancels(
self, mock_confirm, mock_session, mock_list_agents, app_config, settings_store
):
@@ -548,7 +565,10 @@ class TestModifyLLMSettingsAdvanced:
'openhands.cli.settings.LLMSummarizingCondenserConfig',
MockLLMSummarizingCondenserConfig,
)
- @patch('openhands.cli.settings.NoOpCondenserConfig', MockNoOpCondenserConfig)
+ @patch(
+ 'openhands.cli.settings.ConversationWindowCondenserConfig',
+ MockConversationWindowCondenserConfig,
+ )
async def test_modify_llm_settings_advanced_invalid_agent(
self,
mock_print,
@@ -599,7 +619,10 @@ class TestModifyLLMSettingsAdvanced:
'openhands.cli.settings.LLMSummarizingCondenserConfig',
MockLLMSummarizingCondenserConfig,
)
- @patch('openhands.cli.settings.NoOpCondenserConfig', MockNoOpCondenserConfig)
+ @patch(
+ 'openhands.cli.settings.ConversationWindowCondenserConfig',
+ MockConversationWindowCondenserConfig,
+ )
async def test_modify_llm_settings_advanced_user_rejects_save(
self, mock_confirm, mock_session, mock_list_agents, app_config, settings_store
):
diff --git a/tests/unit/test_config.py b/tests/unit/test_config.py
index bb500dcacd..dacae26223 100644
--- a/tests/unit/test_config.py
+++ b/tests/unit/test_config.py
@@ -16,6 +16,7 @@ from openhands.core.config import (
load_openhands_config,
)
from openhands.core.config.condenser_config import (
+ ConversationWindowCondenserConfig,
LLMSummarizingCondenserConfig,
NoOpCondenserConfig,
RecentEventsCondenserConfig,
@@ -680,7 +681,7 @@ def test_agent_config_condenser_with_no_enabled():
"""Test default agent condenser with enable_default_condenser=False."""
config = OpenHandsConfig(enable_default_condenser=False)
agent_config = config.get_agent_config()
- assert isinstance(agent_config.condenser, NoOpCondenserConfig)
+ assert isinstance(agent_config.condenser, ConversationWindowCondenserConfig)
def test_sandbox_volumes_toml(default_config, temp_toml_file):
@@ -907,9 +908,9 @@ def test_default_condenser_behavior_disabled(default_config, temp_toml_file):
default_config.enable_default_condenser = False
load_from_toml(default_config, temp_toml_file)
- # Verify the agent config uses NoOpCondenserConfig
+ # Verify the agent config uses ConversationWindowCondenserConfig
agent_config = default_config.get_agent_config()
- assert isinstance(agent_config.condenser, NoOpCondenserConfig)
+ assert isinstance(agent_config.condenser, ConversationWindowCondenserConfig)
def test_default_condenser_explicit_toml_override(default_config, temp_toml_file):