CrewAI Web Research: Multi-Agent Tutorial

CrewAI lets you orchestrate multiple AI agents working together. This tutorial shows you how to build a research crew with web access.

The Research Crew

• Researcher Agent: Searches and reads web sources
• Analyst Agent: Synthesizes and extracts insights
• Writer Agent: Produces final report

Setup

pip install crewai crewai-tools

Create Tryb Tools

from crewai_tools import BaseTool
import requests

class TrybSearchTool(BaseTool):
    name: str = "Web Search"
    description: str = "Search the web for information"
    
    def _run(self, query: str) -> str:
        response = requests.post(
            "https://api.tryb.dev/v1/search",
            headers={"Authorization": f"Bearer {os.getenv('TRYB_API_KEY')}"},
            json={"query": query, "num_results": 5}
        )
        return str(response.json()["results"])

class TrybReadTool(BaseTool):
    name: str = "Read Webpage"
    description: str = "Read content from a URL"
    
    def _run(self, url: str) -> str:
        response = requests.post(
            "https://api.tryb.dev/v1/read",
            headers={"Authorization": f"Bearer {os.getenv('TRYB_API_KEY')}"},
            json={"url": url}
        )
        return response.json()["data"]["markdown"]

Define Agents

from crewai import Agent, Task, Crew

researcher = Agent(
    role="Research Specialist",
    goal="Find comprehensive information on the given topic",
    backstory="Expert at finding and extracting valuable information from the web",
    tools=[TrybSearchTool(), TrybReadTool()],
    verbose=True
)

analyst = Agent(
    role="Data Analyst",
    goal="Analyze research findings and extract key insights",
    backstory="Skilled at identifying patterns and synthesizing information"
)

writer = Agent(
    role="Content Writer",
    goal="Create clear, engaging content from research",
    backstory="Expert at transforming complex information into readable content"
)

Define Tasks

research_task = Task(
    description="Research {topic} using web search and page reading",
    expected_output="Comprehensive research notes with sources",
    agent=researcher
)

analysis_task = Task(
    description="Analyze research and extract 5 key insights",
    expected_output="Bullet-point list of insights with evidence",
    agent=analyst
)

writing_task = Task(
    description="Write a 500-word article based on the analysis",
    expected_output="Polished article ready for publication",
    agent=writer
)

Run the Crew

crew = Crew(
    agents=[researcher, analyst, writer],
    tasks=[research_task, analysis_task, writing_task],
    verbose=True
)

result = crew.kickoff(inputs={"topic": "Latest AI chip developments"})

Related Tutorials

• LangChain Web Research Agent
• Building Web-Aware AI Agents