File size: 9,747 Bytes
b6ae7b8 | 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 | #!/usr/bin/env python3
"""
Benchmarks for Stack 2.9 - Token Efficiency Tests
Token optimization benchmarks.
"""
import pytest
import sys
from pathlib import Path
from unittest.mock import MagicMock, patch
# Add stack_cli to path
sys.path.insert(0, str(Path(__file__).parent.parent / "stack_cli"))
from stack_cli.agent import StackAgent, create_agent, AgentResponse
class TestTokenUsage:
"""Test token usage patterns."""
def test_response_token_efficiency(self):
"""Test response token efficiency."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.tools.get_tool') as mock_get_tool:
mock_tool = MagicMock(return_value={"success": True, "content": "x"})
mock_get_tool.return_value = mock_tool
agent = StackAgent()
response = agent.process("read test.py")
# Response should have content
assert response.content is not None
assert len(response.content) > 0
def test_context_truncation(self):
"""Test context truncation."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.context.Path') as mock_path:
with patch.object(Path, 'exists', return_value=False):
from stack_cli.context import ContextManager
cm = ContextManager("/tmp")
# Context should be generated
context = cm.get_workspace_context()
# Should be formatted string
assert isinstance(context, str)
class TestPromptEfficiency:
"""Test prompt efficiency."""
def test_intent_parsing_tokens(self):
"""Test intent parsing token usage."""
from stack_cli.agent import QueryUnderstanding
qu = QueryUnderstanding()
# Parse should be efficient
result = qu.parse("read test.py")
# Result should have required fields
assert "intent" in result
assert "confidence" in result
assert result["intent"] == "file_read"
def test_tool_selection_tokens(self):
"""Test tool selection token usage."""
from stack_cli.agent import ToolSelector
ts = ToolSelector()
# Selection should be minimal
tools = ts.select("file_read", {})
# Should return list of tools
assert isinstance(tools)
assert len(tools) > 0
class TestResponseEfficiency:
"""Test response generation efficiency."""
def test_response_generation_size(self):
"""Test response generation output size."""
with patch('stack_cli.context.create_context_manager'):
from stack_cli.agent import ResponseGenerator, ToolCall
rg = ResponseGenerator()
tool_calls = [
ToolCall(
tool_name="read",
arguments={"path": "test.py"},
result={"success": True, "content": "test content"},
success=True
)
]
response = rg.generate(tool_calls, "file_read", {})
# Should produce reasonable output
assert isinstance(response, str)
assert len(response) > 0
# Should not be excessively long
assert len(response) < 10000
def test_clarification_efficiency(self):
"""Test clarification generation efficiency."""
with patch('stack_cli.context.create_context_manager'):
from stack_cli.agent import ResponseGenerator
rg = ResponseGenerator()
clarification = rg.generate_clarification("Which file?")
# Should be concise
assert isinstance(clarification, str)
assert len(clarification) < 200
class TestContextTokenEfficiency:
"""Test context token efficiency."""
def test_context_summary_size(self):
"""Test context summary size."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.context.Path') as mock_path:
with patch.object(Path, 'exists', return_value=False):
from stack_cli.context import ContextManager
cm = ContextManager("/tmp")
summary = cm.get_context_summary()
# Should be JSON-serializable dict
import json
serialized = json.dumps(summary)
# Should be reasonable size
assert len(serialized) < 10000
def test_workspace_context_size(self):
"""Test workspace context size."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.context.Path') as mock_path:
with patch.object(Path, 'exists', return_value=False):
from stack_cli.context import ContextManager
cm = ContextManager("/tmp")
context = cm.get_workspace_context()
# Should be reasonable size
assert len(context) < 10000
class TestToolSchemasEfficiency:
"""Test tool schemas token efficiency."""
def test_schemas_compactness(self):
"""Test schemas are compact."""
from stack_cli.tools import get_tool_schemas
schemas = get_tool_schemas()
import json
serialized = json.dumps(schemas)
# Should be reasonable size
assert len(serialized) < 50000
def test_schema_required_fields(self):
"""Test schemas have required fields only."""
from stack_cli.tools import get_tool_schemas
schemas = get_tool_schemas()
for schema in schemas:
# Should have minimal required fields
assert "name" in schema
assert "description" in schema
assert "parameters" in schema
# Parameters should be minimal
params = schema["parameters"]
assert "type" in params
assert "properties" in params
class TestConversationEfficiency:
"""Test conversation token efficiency."""
def test_history_truncation(self):
"""Test conversation history truncation."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.tools.get_tool') as mock_get_tool:
mock_tool = MagicMock(return_value={"success": True})
mock_get_tool.return_value = mock_tool
agent = StackAgent()
# Add many conversations
for i in range(50):
agent.process(f"query {i}")
# History may be truncated
history_len = len(agent.conversation_history)
# Should not grow unbounded
assert history_len <= 100
def test_summary_efficiency(self):
"""Test summary efficiency."""
with patch('stack_cli.context.create_context_manager'):
agent = StackAgent()
session = agent.context_manager.session
# Add some data
for i in range(10):
session.add_message("user", f"message {i}")
summary = session.get_summary()
# Summary should be compact
import json
serialized = json.dumps(summary)
assert len(serialized) < 1000
class TestTokenOptimization:
"""Test token optimization strategies."""
def test_response_capping(self):
"""Test response content capping."""
with patch('stack_cli.context.create_context_manager'):
from stack_cli.agent import ResponseGenerator, ToolCall
rg = ResponseGenerator()
# Long content should be capped
long_content = "x" * 10000
tool_calls = [
ToolCall(
tool_name="read",
arguments={"path": "test.py"},
result={"success": True, "content": long_content},
success=True
)
]
response = rg.generate(tool_calls, "file_read", {})
# Response should be capped
assert len(response) < 15000
def test_context_truncation_strategy(self):
"""Test context truncation strategy."""
with patch('stack_cli.context.create_context_manager'):
with patch('stack_cli.context.Path') as mock_path:
with patch.object(Path, 'exists', return_value=False):
from stack_cli.context import ContextManager
cm = ContextManager("/tmp")
# With no projects, context should be minimal
context = cm.get_workspace_context()
# Should be concise
lines = context.split('\n')
# Should not have excessive lines
assert len(lines) < 100
if __name__ == "__main__":
pytest.main([__file__, "-v"])
|