llama-index-tools-agentbay 0.1.0

AgentBay tools integration for LlamaIndex - provides browser automation, file operations, and command execution capabilities

LlamaIndex Tools Integration: AgentBay

This package provides AgentBay tools integration for LlamaIndex, enabling browser automation, file operations, and command execution capabilities in your LlamaIndex applications.

Installation

pip install llama-index-tools-agentbay

Requirements

• Python >= 3.8
• llama-index-core >= 0.12.0
• wuying-agentbay-sdk >= 0.9.0
• AgentBay API key (get it from AgentBay Console)

Quick Start

import os
from llama_index.llms.dashscope import DashScope
from llama_index.core.agent import ReActAgent
from llama_index.tools.agentbay import create_browser_agent_tools

# Create AgentBay browser tools (automatically handles all setup)
browser_tools = create_browser_agent_tools()

# Create LLM
llm = DashScope(
    model_name="qwen-plus",
    api_key=os.getenv("DASHSCOPE_API_KEY")
)

# Create Agent with AgentBay tools
agent = ReActAgent.from_tools(browser_tools, llm=llm, verbose=True)

# Use the agent
response = agent.chat("Please take a screenshot of https://www.example.com")
print(response)

Features

Two Simple Tool Sets

For Browser Automation:

• create_browser_agent_tools() - Complete browser automation toolkit
  • Screenshot capture
  • Browser endpoint access (for Playwright)
  • File operations
  • Command execution

For Code Execution:

• create_code_agent_tools() - Complete code execution toolkit
  • Python code execution
  • JavaScript code execution
  • File operations (read, write, list, download)
  • Command execution

Usage Examples

Browser Automation

from llama_index.tools.agentbay import create_browser_agent_tools
from llama_index.core.agent import ReActAgent

# Create browser tools (screenshot, file ops, commands)
tools = create_browser_agent_tools()

# Create agent
agent = ReActAgent.from_tools(tools, llm=llm)

# Use the agent
response = agent.chat("Take a screenshot of https://github.com")

Code Execution

from llama_index.tools.agentbay import create_code_agent_tools
from llama_index.core.agent import ReActAgent

# Create code execution tools (Python, JavaScript, file ops, commands)
tools = create_code_agent_tools()

# Create agent
agent = ReActAgent.from_tools(tools, llm=llm)

# Use the agent to run code
response = agent.chat("""
Create a CSV file with sample data and analyze it with pandas:
1. Create a CSV with sales data (product, quantity, price)
2. Read it with pandas
3. Calculate total revenue
4. Show the results
""")

Benefits:

• Direct code execution (no file upload needed)
• Supports Python and JavaScript
• Includes file operations for saving results
• Faster and more reliable

Choosing the Right Tool Set

Simple decision:

• Need browser automation? → create_browser_agent_tools()
• Need to run code (Python/JS)? → create_code_agent_tools()

Both tool sets include file operations and command execution, so you get everything you need in one function call.

Environment Variables

Set your API keys as environment variables:

export AGENTBAY_API_KEY="your-agentbay-api-key"
export DASHSCOPE_API_KEY="your-dashscope-api-key"  # For DashScope LLM

Advanced Features

RAG Integration for Insights

Automatically extract and index insights from code execution results:

from llama_index.tools.agentbay import create_code_agent_tools, create_rag_manager

# Create tools and RAG manager
tools = create_code_agent_tools()
rag_manager = create_rag_manager()

# Use with your agent to execute code and store results
# The RAG manager can index execution results for later querying

# Query the RAG index
answer = rag_manager.query("What insights did we discover?")
print(answer)

Complete 4-Stage Demo: See /examples/ for full workflow examples.

API Reference

Main Functions

create_browser_agent_tools()

• Returns: List of tools for browser automation
• Includes: screenshot, browser info, file operations, commands
• Use when: You need to automate browser tasks

create_code_agent_tools()

• Returns: List of tools for code execution
• Includes: Python/JS execution, file operations, commands
• Use when: You need to run code or process data

Both functions automatically handle session management and cleanup.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License

Links

• AgentBay Console
• AgentBay SDK Documentation
• LlamaIndex Documentation

Support

For issues and questions:

• AgentBay SDK: GitHub Issues
• LlamaIndex: GitHub Issues