LangChain Agent Config

A Python library for managing multiple LangChain model configurations with validation and automatic agent creation.

Features

• ✅ Multi-provider support: Gemini (Google), OpenAI (GPT), and Anthropic (Claude)
• ✅ Robust validation: Pydantic-based validation with custom validators
• ✅ Environment-based configuration: Support for multi-line .env files
• ✅ Automatic agent creation: Creates LangChain agents from configuration
• ✅ Type-safe: Full type hints and Pydantic models
• ✅ Well tested: 43+ tests with pytest

Installation

pip install langchain-agent-config

Quick Start

1. Create a .env file in your project root:

MODELS=[
    {
        "name":"gemini",
        "model":"gemini-2.0-flash-exp",
        "temperature":0.7,
        "key":"your-google-api-key",
        "max_tokens":2048
    },
    {
        "name":"gpt",
        "model":"gpt-4",
        "temperature":0.5,
        "key":"your-openai-api-key",
        "max_tokens":4096
    }
]

2. Use in your code:

from langchain_agent_config import ModelsEnv, read_env_file
from pathlib import Path
import json

# Load configuration
env_path = Path(".env")
env_vars = read_env_file(env_path)
raw_models = json.loads(env_vars["MODELS"])

# Validate and create models
models_env = ModelsEnv(models=raw_models)

# Create agents
agents = models_env.create_agents()

# Use the agents
response = agents['gemini'].invoke("Hello!")
response2 = agents['gpt'].invoke("Hello!")

Configuration

Model Configuration Fields

• name (str): Model identifier (must contain 'gemini', 'gpt'/'openai', or 'anthropic'/'claude')
• model (str): Specific model name from the provider
• temperature (float): Value between 0.0 and 1.0
• key (str): API key for the model provider
• max_tokens (int): Maximum number of tokens (must be positive)

Supported Providers

• Gemini: Models with "gemini" in the name
• OpenAI/GPT: Models with "gpt" or "openai" in the name
• Anthropic/Claude: Models with "anthropic" or "claude" in the name

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Run with coverage
pytest --cov=backend.model

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.