Lesson 3 of 6~2 min read

Lesson 2 — Setting Up Your Tools

You need two things: a way for the AI to search the web, and a way for it to read web pages.

You need two things: a way for the AI to search the web, and a way for it to read web pages.

Install dependencies:

pip install anthropic tavily-python requests beautifulsoup4

Get your API keys:

Anthropic: console.anthropic.com (you may already have this)

Tavily: tavily.com (free tier gives 1,000 searches/month — plenty)

Create a `.env` file:

ANTHROPIC_API_KEY=your_key_here

TAVILY_API_KEY=your_key_here

Define your tools in Python:

import os

import anthropic

from tavily import TavilyClient

import requests

from bs4 import BeautifulSoup

anthropic_client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])

tavily_client = TavilyClient(api_key=os.environ["TAVILY_API_KEY"])

def search_web(query: str, max_results: int = 5) -> list:

"""Search the web and return results with titles, URLs, and snippets."""

results = tavily_client.search(query, max_results=max_results)

return results["results"]

def read_page(url: str) -> str:

"""Fetch a webpage and return its text content."""

try:

    resp = requests.get(url, timeout=10, headers={"User-Agent": "Mozilla/5.0"})

    soup = BeautifulSoup(resp.text, "html.parser")

    # Remove nav, footer, scripts

    for tag in soup(["script", "style", "nav", "footer", "header"]):

        tag.decompose()

    text = soup.get_text(separator="\n", strip=True)

    # Limit to first 3000 chars to stay within context

    return text[:3000]

except Exception as e:

    return f"Error reading page: {e}"

These two functions are your agent's eyes and hands. The AI will call them as it sees fit.