🛠️ Agent Tools

A community-driven collection of tools for LangChain/LangGraph agents. Build smarter agents by plugging in ready-to-use tools or contribute your own.

Quick Start

# Clone the repository
git clone https://github.com/yourusername/agent-tools.git
cd agent-tools

# Create virtual environment with uv
uv venv
source .venv/bin/activate

# Install dependencies
uv pip install -r todo/requirements.txt

# Set up environment variables
cp .env.example .env
# Add your API keys to .env

Available Tools

Category	Tools	Description	Status
Bash & Shell	bash, glob, grep	Execute shell commands and perform advanced searches	✅ Ready
File System	read_file, write_file, edit_file	Robust file manipulation with safety checks	✅ Ready
Todo Manager	create_todo, list_todos, update_todo	Track progress during multi-step agent tasks	✅ Ready
Interactive	ask_question	Human-in-the-loop interaction for clarifying questions	✅ Ready
Subagents	subagent	Delegate complex tasks to specialized sub-agents	✅ Ready
Skills	-	Extensible framework for adding custom capabilities	🚧 Beta

Usage

Using the Tools

The tools are designed to be used with LangChain or LangGraph agents. Here's how to initialize a basic agent with our tool collection:

from base.models import ToolCollection
from bash.tools import BASH_TOOLS
from file_system.tools import FILE_SYSTEM_TOOLS
from todo.tools import TODO_TOOLS

# Combine toolsets
all_tools = BASH_TOOLS + FILE_SYSTEM_TOOLS + TODO_TOOLS

# Or use the ToolCollection to load everything
# collection = ToolCollection().load_all()

Example: File Editing

# The agent can intelligently edit files instead of rewriting them
edit_file(
    file_path="/path/to/script.py",
    old_content="def old_func():\n    pass",
    new_content="def new_func():\n    print('Hello World')",
)

Contributing

We welcome contributions! Here's how to add your own tools:

Adding a New Tool

1. Fork & Clone

   git clone https://github.com/yourusername/agent-tools.git
   cd agent-tools

2. Create a new tool directory

   mkdir my_tool
   cd my_tool

3. Structure your tool

   my_tool/
   ├── __init__.py
   ├── tools.py          # Your tool implementations
   ├── prompt.md         # System prompt for the agent
   └── requirements.txt  # Tool-specific dependencies
   

4. Implement your tools in tools.py:

   from langchain_core.tools import tool
   from typing import Annotated
   
   @tool
   def my_awesome_tool(
       param: Annotated[str, "Description of the parameter"]
   ) -> str:
       """Clear docstring explaining what the tool does.
       
       Include:
       - When to use this tool
       - What it returns
       - Example usage
       """
       # Your implementation
       return "result"
   
   MY_TOOLS = [my_awesome_tool]

5. Write a system prompt in prompt.md explaining how the agent should use your tools

6. Submit a Pull Request

Tool Guidelines

• ✅ Use clear, descriptive tool names
• ✅ Write comprehensive docstrings (agents read these!)
• ✅ Use Annotated types for parameter descriptions
• ✅ Include a prompt.md with usage guidelines
• ✅ Add tool-specific dependencies to a local requirements.txt
• ✅ Include example usage in your PR description

Pull Request Process

1. Ensure your tool works with the latest LangChain/LangGraph
2. Update the README table with your new tool
3. Add any new dependencies
4. Submit PR with a clear description and example usage

Project Structure

agent-tools/
├── agent.py            # Example agent implementation
├── base/               # Shared base classes and utilities
├── bash/               # Shell execution & search tools (grep, glob)
├── file_system/        # File I/O and targeted editing tools
├── interactive/        # Human-in-the-loop interaction tools
├── skills/             # Reusable agent skill modules
├── subagents/          # Sub-agent delegation tools
└── todo/               # Multi-step task tracking tools

Environment Variables

Create a .env file with your API keys:

GOOGLE_API_KEY=your_google_api_key
OPENAI_API_KEY=your_openai_api_key
# Add other provider keys as needed

Requirements

• Python 3.10+
• LangChain 1.0+
• LangGraph 1.0+

License

MIT License - see LICENSE for details.

Show Your Support

Give a ⭐ if this project helped you build better agents!

---

Built with ❤️ by the community