Skip to main content
Unlisted page
This page is unlisted. Search engines will not index it, and only users having a direct link can access it.

Build Custom AI Agents with Enigma

Beta

Enigma's remote Model Context Protocol (MCP) server is currently in beta.

Integrate Enigma directly into your custom AI agents and applications. Works with any framework - OpenAI Agents, LangChain, CrewAI, AutoGen, and more.

Prerequisites​

  • An Enigma API key, which you can find on the homepage of the Enigma Console
  • Python development environment

Quick Start with OpenAI Agents​

Method 1: API Key Authentication (Recommended)

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerStreamableHttp

async def main():
    async with MCPServerStreamableHttp(
        params={
            "url": "https://mcp.enigma.com/http-key",
            "headers": {"x-api-key": "your_api_key"}
        }
    ) as server:
        await server.connect()
        
        agent = Agent(
            name="Business Intelligence Agent",
            mcp_servers=[server]
        )
        
        result = Runner.run_streamed(
            starting_agent=agent,
            input="Analyze Tacombi's financial performance and expansion strategy"
        )

asyncio.run(main())

Method 2: OAuth Authentication

import asyncio
from agents.mcp import MCPServerStreamableHttp

async def main():
    async with MCPServerStreamableHttp(
        params={"url": "https://mcp.enigma.com/http"}
    ) as server:
        await server.connect()
        # OAuth flow will open automatically
        
        # Your agent code here...

asyncio.run(main())

Framework Examples​

LangChain Integration

from langchain_openai import ChatOpenAI
from langchain_core.tools import BaseTool
from mcp_client import MCPClient

class EnigmaBusinessSearch(BaseTool):
    name = "search_business"
    description = "Search for business information"
    
    def _run(self, business_name: str) -> str:
        # MCP client integration
        client = MCPClient("https://mcp.enigma.com/http-key", 
                          headers={"x-api-key": "your_api_key"})
        return client.call_tool("search_business", {"name": business_name})

# Use in your LangChain agents
llm = ChatOpenAI()
tools = [EnigmaBusinessSearch()]

CrewAI Integration

from crewai import Agent, Task, Crew
from crewai_tools import BaseTool

class EnigmaTool(BaseTool):
    name: str = "Enigma Business Intelligence"
    description: str = "Access comprehensive business data and analytics"
    
    def _run(self, business_name: str) -> str:
        # Your MCP integration here
        pass

business_analyst = Agent(
    role='Business Analyst',
    goal='Analyze business performance and competitive landscape',
    backstory='Expert in business intelligence and market analysis',
    tools=[EnigmaTool()],
    verbose=True
)

Complete Examples Repository​

Get started quickly: Clone our comprehensive examples repository with ready-to-use implementations:

git clone https://github.com/enigma-io/enigma-mcp-examples
cd enigma-mcp-examples
pip install -r requirements.txt

Available examples:

  • OpenAI Agents with business intelligence workflows
  • LangChain integration for market research
  • CrewAI multi-agent business analysis teams
  • AutoGen collaborative business planning agents
  • Custom REST API wrapper for any framework

Authentication Options​

API Key (Recommended for Production)

  • More reliable for automated systems
  • No browser interaction required
  • Better for CI/CD and serverless environments
headers = {"x-api-key": "your_enigma_api_key"}

OAuth (Good for Development)

  • Browser-based authentication flow
  • Uses same credentials as Enigma Console
  • Better for local development and testing

Common Use Cases​

Due Diligence Agents

# Automate comprehensive business investigations
tools = ["search_business", "get_brand_legal_entities", 
         "search_negative_news", "search_gov_archive"]

Market Research Agents

# Competitive analysis and market sizing
tools = ["generate_brands_segment", "get_brand_locations",
         "get_brand_card_analytics"]

Compliance Monitoring Agents

# Ongoing business monitoring and risk assessment  
tools = ["search_kyb", "search_gov_archive", "search_negative_news"]

Troubleshooting​

API Key authentication failing?

  • Verify your API key at Enigma Console
  • Ensure the x-api-key header is properly set

OAuth flow not working?

  • Check that your development environment can open browser windows
  • Verify you're logged into Enigma Console in your default browser

MCP connection issues?

  • Ensure your framework supports HTTP MCP servers
  • Check network connectivity to mcp.enigma.com

Need help with integration?

Resources​

Last updated on