langchain-local-dev-loop（LangChain 本地 DEV loop）

LangChain Local Dev Loop

Overview

Configure a rapid local development workflow for LangChain applications with testing, debugging, and hot reload capabilities.

Prerequisites

• Completed langchain-install-auth setup
• Python 3.9+ with virtual environment
• pytest and related testing tools
• IDE with Python support (VS Code recommended)

Instructions

Step 1: Set Up Project Structure

my-langchain-app/
├── src/
│   ├── __init__.py
│   ├── chains/
│   │   └── __init__.py
│   ├── agents/
│   │   └── __init__.py
│   └── prompts/
│       └── __init__.py
├── tests/
│   ├── __init__.py
│   ├── conftest.py
│   └── test_chains.py
├── .env
├── .env.example
├── pyproject.toml
└── README.md

Step 2: Configure Testing

# tests/conftest.py
import pytest
from unittest.mock import MagicMock
from langchain_core.messages import AIMessage

@pytest.fixture
def mock_llm():
    """Mock LLM for unit tests without API calls."""
    mock = MagicMock()
    mock.invoke.return_value = AIMessage(content="Mocked response")
    return mock

@pytest.fixture
def sample_prompt():
    """Sample prompt for testing."""
    from langchain_core.prompts import ChatPromptTemplate
    return ChatPromptTemplate.from_template("Test: {input}")

Step 3: Create Test File

# tests/test_chains.py
def test_chain_construction(mock_llm, sample_prompt):
    """Test that chain can be constructed."""
    from langchain_core.output_parsers import StrOutputParser

    chain = sample_prompt | mock_llm | StrOutputParser()
    assert chain is not None

def test_chain_invoke(mock_llm, sample_prompt):
    """Test chain invocation with mock."""
    from langchain_core.output_parsers import StrOutputParser

    chain = sample_prompt | mock_llm | StrOutputParser()
    result = chain.invoke({"input": "test"})
    assert result == "Mocked response"

Step 4: Set Up Development Tools

# pyproject.toml
[project]
name = "my-langchain-app"
version = "0.1.0"
requires-python = ">=3.9"
dependencies = [
    "langchain>=0.3.0",
    "langchain-openai>=0.2.0",
    "python-dotenv>=1.0.0",
]

[project.optional-dependencies]
dev = [
    "pytest>=8.0.0",
    "pytest-asyncio>=0.23.0",
    "pytest-cov>=4.0.0",
    "ruff>=0.1.0",
    "mypy>=1.0.0",
]

[tool.pytest.ini_options]
asyncio_mode = "auto"
testpaths = ["tests"]

[tool.ruff]
line-length = 100

Output

• Organized project structure with separation of concerns
• pytest configuration with fixtures for mocking LLMs
• Development dependencies configured
• Ready for rapid iteration

Error Handling

Examples

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=src --cov-report=html

# Run specific test
pytest tests/test_chains.py::test_chain_invoke -v

# Watch mode (requires pytest-watch)
ptw

Integration Test Example

# tests/test_integration.py
import pytest
from dotenv import load_dotenv

load_dotenv()

@pytest.mark.integration
def test_real_llm_call():
    """Integration test with real LLM (requires API key)."""
    from langchain_openai import ChatOpenAI

    llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
    response = llm.invoke("Say 'test passed'")
    assert "test" in response.content.lower()

Resources

• pytest Documentation
• LangChain Testing Guide
• python-dotenv

Next Steps

Proceed to langchain-sdk-patterns for production-ready code patterns.