Built-in Tools: No Coding Required

You don't always have to write your own tools. LangChain comes with a massive library of Community Tools for common tasks. Using these saves you hours of API integration work.

1. Tavily (Search for AI)

Tavily is a search engine designed specifically for Agents. It doesn't return ads; it returns clean text that is easy for LLMs to read.

from langchain_community.tools.tavily_search import TavilySearchResults

# 1. Setup (Requires TAVILY_API_KEY in .env)
search = TavilySearchResults(k=3)

# 2. Use it directly or in an agent
results = search.invoke({"query": "What is the capital of France?"})

---

2. PythonREPL (The Sandbox)

This tool allows the agent to write and execute its own Python code to solve math problems or create charts.

• DANGER: Never use this in production without a secure Docker sandbox! (Module 13 of Agentic Course).

---

3. Wikipedia API

A classic tool for historical or general knowledge.

from langchain_community.tools import WikipediaQueryRun
from langchain_community.utilities import WikipediaAPIWrapper

wikipedia = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())

---

4. Visualizing the Toolset

Tool	Capability	Best Use Case
Tavily	Real-time Web	Current events, price checks.
PythonREPL	Logic / Calculation	Complex math, data analysis.
DuckDuckGo	Search	Private, anonymous web search.
YouTube	Transcript Retrieval	Summarizing videos.

---

5. Engineering Tip: Tool Persistence

Some tools require long-running connections (like a SQL tool). Use the load_tools utility to manage their lifecycle efficiently.

from langchain.agents import load_tools
tools = load_tools(["wikipedia", "llm-math"], llm=model)

---

Key Takeaways

• LangChain Community provides hundreds of ready-to-use tools.
• Tavily is the current standard for agentic web search.
• PythonREPL gives the agent "Infinite" logical capability via code.
• Always check the Security of a tool before giving it to an autonomous agent.