StitchLab Agent Core

A powerful agent core application built with modern Python frameworks for AI agent development.

Features

• FastAPI-based runtime environment
• Support for multiple AI backends via LiteLLM
• Langfuse integration for monitoring and observability
• MCP (Model Context Protocol) support
• Pydantic-based schema validation
• Asyncio-native async/await support

Installation

Install from PyPI:

pip install stitchlab-agentcore

Requirements

• Python 3.11 or higher

Quick Start

Basic Example

Here's a complete example of how to use the StitchLab Agent Core library:

import os
from dotenv import load_dotenv
load_dotenv()

from stitchlab_agentcore.runtime.factory import AgentFactory
from stitchlab_agentcore.runtime.app import StitchLabAgentCoreApp
from stitchlab_agentcore.config import GlobalConfig, BaseSettings

from typing import Optional
from strands import Agent, tool
import logging

logger = logging.getLogger(__name__)


### ================ Implement your Strands Agent ================ 

class GlobalSettings(BaseSettings):
    APP_NAME: str = 'Your Strands Agent App Name'
    VERBOSE: bool = os.getenv('VERBOSE', 'True').lower() in ('true', '1', 'yes')
    DEBUG: bool = os.getenv('DEBUG', 'False').lower() in ('true', '1', 'yes')
    VERIFY_CERTIFICATE: bool = os.getenv('VERIFY_CERTIFICATE', 'True').lower() in ('true', '1', 'yes')

    MCP_URL: str = os.getenv('MCP_URL')
    MCP_TOOLS: list[str] = [tool.strip() for tool in os.getenv('MCP_TOOLS', '-').split(',')]
    
    MODEL_ID: str = os.getenv('BEDROCK_MODEL_ID', 'bedrock/anthropic.claude-3-5-sonnet-20240620-v1:0')
    MEMORY_ID: str = os.getenv('BEDROCK_AGENTCORE_MEMORY_ID')
    BEDROCK_REGION: str = os.getenv('BEDROCK_REGION')
    
    BEDROCK_GUARDRAIL_TRACE: str = "disabled"
    BEDROCK_GUARDRAIL_ID: Optional[str] = None
    BEDROCK_GUARDRAIL_VER: Optional[str] = None

    LANGFUSE_PUBLIC_KEY: Optional[str] = os.getenv("LANGFUSE_PUBLIC_KEY")
    LANGFUSE_SECRET_KEY: Optional[str] = os.getenv("LANGFUSE_SECRET_KEY")
    LANGFUSE_HOST: Optional[str] = os.getenv("LANGFUSE_HOST")


class AppConfig(GlobalConfig[GlobalSettings]):
    pass

CONFIG = AppConfig(GlobalSettings())


@tool
def subtract(a: int, b: int) -> int:
    """Calculate the difference between two numbers"""
    return a - b

@tool
def multiply(a: int, b: int) -> int:
    """Calculate the product of two numbers"""
    return a * b


SYSTEM_PROMPT = """
You are a good friend.. Be nice
"""


TOOLS = [
    subtract,
    multiply
]

### ================ End of Implementation ================ 


AGENT_FACTORY = AgentFactory(
    system_prompt=SYSTEM_PROMPT,
    local_tools=TOOLS,
    config=CONFIG
)

async def create_agent(actor_id: str, session_id: str, trace_attributes: Optional[dict] = None) -> Optional[Agent]:
    return await AGENT_FACTORY.create_agent(actor_id=actor_id, session_id=session_id, trace_attributes=trace_attributes)


app = StitchLabAgentCoreApp(debug=True).initialize()

@app.agent_entrypoint(create_agent)
async def agent_invocation(payload):
    pass


if __name__ == "__main__":
    CONFIG.logger.info(f"Starting {CONFIG.settings.APP_NAME} on FastAPI server...")
    app.run()

Configuration

Configure the application using environment variables or .env file:

# Add your configuration here

Architecture

The StitchLab Agent Core provides:

• Runtime: FastAPI-based application runtime for agents
• Schema: Pydantic models for data validation
• Config: Centralized configuration management
• Assets: Built-in resources and caching

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributing

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

Support

For issues, questions, or contributions, please visit the GitHub repository.