Implementation:CrewAIInc CrewAI CrewAI Tools Library

This is a Wrapper Doc for the crewai-tools library.

Overview

Curated library of 60+ pre-built agent tools for search, scraping, file I/O, databases, and AI services provided by the crewai-tools package.

Source Reference

• Repository: crewAIInc/crewAI
• File: lib/crewai-tools/src/crewai_tools/__init__.py
• Lines: L197-292

Import

from crewai_tools import SerperDevTool, FileReadTool, PDFSearchTool, ...

Tool Categories

Search Tools

Tool	Provider	Description
SerperDevTool	Serper.dev	Google search results via the Serper API
TavilySearchTool	Tavily	AI-optimized search engine with structured results
BraveSearchTool	Brave	Privacy-focused web search
EXASearchTool	EXA	Neural search engine for semantic queries

Scraping Tools

Tool	Approach	Description
ScrapeWebsiteTool	HTTP	Simple HTTP-based web page content extraction
FirecrawlScrapeWebsiteTool	Managed service	Scraping via the Firecrawl managed service
SeleniumScrapingTool	Browser automation	JavaScript-rendered page scraping via Selenium

File I/O Tools

Tool	Description
FileReadTool	Read the contents of a file from the local filesystem
FileWriterTool	Write content to a file on the local filesystem
PDFSearchTool	Search and extract content from PDF documents
DirectoryReadTool	List files and directories within a path

Database Tools

Tool	Database	Description
MySQLSearchTool	MySQL	Execute queries against a MySQL database
PGSearchTool	PostgreSQL	Execute queries against a PostgreSQL database
SnowflakeSearchTool	Snowflake	Query a Snowflake data warehouse

Code Tools

Tool	Description
CodeInterpreterTool	Execute Python code in a sandboxed environment and return results

RAG Tools

Tool	Description
CSVSearchTool	Semantic search over CSV file contents
JSONSearchTool	Semantic search over JSON file contents
MDXSearchTool	Semantic search over MDX/Markdown documents

Example

Importing and Configuring Common Tools

import os
from crewai import Agent
from crewai_tools import (
    SerperDevTool,
    FileReadTool,
    PDFSearchTool,
    ScrapeWebsiteTool,
)

# Configure API keys via environment variables
os.environ["SERPER_API_KEY"] = "your-serper-api-key"

# Instantiate built-in tools
search_tool = SerperDevTool()
file_reader = FileReadTool()
pdf_search = PDFSearchTool(pdf="reports/annual_report.pdf")
scraper = ScrapeWebsiteTool()

# Assign tools to an agent
research_agent = Agent(
    role="Research Analyst",
    goal="Find and analyze information from multiple sources",
    backstory="An experienced analyst skilled at gathering data.",
    tools=[search_tool, file_reader, pdf_search, scraper],
)

Using Database Tools

from crewai_tools import PGSearchTool

# Configure a PostgreSQL search tool
pg_tool = PGSearchTool(
    db_uri="postgresql://user:password@localhost:5432/mydb",
)

# The agent can now query the database using natural language
db_agent = Agent(
    role="Data Analyst",
    goal="Query the database to answer business questions",
    backstory="An analyst with deep SQL expertise.",
    tools=[pg_tool],
)

Installation

# Install the tools package separately
pip install crewai-tools

# Or install crewai with tools extras
pip install crewai[tools]

Principle Link

Principle:CrewAIInc_CrewAI_Built_In_Tool_Selection

See Also

• Implementation:CrewAIInc_CrewAI_BaseTool_Schema -- The base class all built-in tools extend
• Implementation:CrewAIInc_CrewAI_Tool_Decorator_And_Classes -- The patterns used to implement these tools
• Implementation:CrewAIInc_CrewAI_Tool_Assignment_Config -- How to assign built-in tools to agents and tasks
• Environment:CrewAIInc_CrewAI_Optional_Provider_Dependencies