CrewAI workflow ideas
SDR agent crew
A Researcher finds leads, a Writer drafts emails, and a Sender triggers outreach.
Market mapping agent
Map all businesses of a given type in a territory and summarise market saturation.
Lead scoring crew
Find 500 leads, score by website presence, ratings, and social gaps, then return the top 20.
Partnership research
Find potential local business partners and produce a shortlist with rationale.
Example CrewAI tool
from crewai import Agent
from crewai.tools import tool
import requests
@tool("Margo Lead Finder")
def find_leads(keyword_and_location: str) -> str:
parts = keyword_and_location.split(" in ", 1)
keyword = parts[0]
location = parts[1] if len(parts) > 1 else ""
response = requests.post(
"https://www.margoleads.io/api/v1/leads/generate",
headers={"Authorization": "Bearer YOUR_API_KEY"},
json={"keyword": keyword, "location": location, "limit": 25}
)
leads = response.json().get("leads", [])
return "\n".join([f"{lead['businessName']} | {lead['email']}" for lead in leads[:10]])
researcher = Agent(
role="Lead Researcher",
goal="Find verified local business leads matching the user's ICP",
tools=[find_leads],
verbose=True
)Frequently asked questions
What's the difference between LangChain and CrewAI integration?
They use the same REST API. CrewAI typically registers Margo through a @tool decorator while LangChain uses a Tool wrapper.
Can multiple agents call Margo?
Yes. Multiple agents can call Margo with the same API key. Rate limits depend on your plan.
Is there a CrewAI template?
A template is planned. The current API is simple enough to register as a CrewAI tool directly.