Agentic Tools Generator

The Tools Generator dynamically creates GLLM-Tools-compatible tools by fetching connector schemas from the Connector API. This approach ensures that your tools are always up-to-date with the latest connector specifications and capabilities.

Key Features

• Dynamic Tool Generation: Automatically generates tools based on live connector schemas

• LangChain Integration: Produces BaseTool compatible tools for seamless integration

• Multiple Connector Support: Supports various popular services and platforms

• API-Driven: Fetches the latest connector specifications from GL Connectors API

Installation

pip

pip install gl-connectors-binary

poetry

uv

Configuration

Before using the GL Connectors Tools Generator, you need to configure your environment variables:

# Required: Your GL Connectors API key
export API_KEY="your`-api-key-here"

# Optional: Custom GL Connectors API base URL (defaults to https://connector.gdplabs.id)
export API_BASE_URL="https://connector.gdplabs.id"

Usage

Basic Usage

Here's how to generate tools for a specific GL Connectors connector:

from gl_connectors import GLConnectorToolGenerator
from langchain_core.tools import BaseTool
import os

def create_tools(connector_name: str) -> list[BaseTool]:
    """Create tools for connector.

    Args:
        app_name: Name of the connector.

    Returns:
        List of tools generated by GLConnectorToolGenerator.
    """    
    generator = GLConnectorToolGenerator(
        api_base_url=os.getenv("API_BASE_URL", "https://connector.gdplabs.id"),
        api_key=os.getenv("API_KEY", ""),
        app_name=connector_name,
    )
    return generator.generate_tools()

Example: Creating GitHub Tools

# Generate GitHub connector tools
github_tools = create_tools("github")

# Use the tools with your AI agent
for tool in github_tools:
    print(f"Tool: {tool.name}")
    print(f"Description: {tool.description}")
    print("---")

API Reference

GLConnectorToolGenerator

The main class for generating Connector integration tools.

Constructor Parameters

Parameter	Type	Description	Required
api_base_url	str	Base URL for the Connector API	No (defaults to https://connector.gdplabs.id)
api_key	str	Your Connector API Client key	Yes
app_name	str	Name of the connector module	Yes

Methods

generate_tools() -> list[BaseTool]

Generates and returns a list of LangChain-compatible tools based on the connector schema.

Returns:

• list[BaseTool]: List of generated tools ready for use with LangChain agents

Raises:

• ValueError: If the app_name is not supported

• APIError: If there's an issue connecting to the Connector API

create_tools Function

A convenience function that wraps the GLConnectorToolGenerator with validation.

Parameters

Parameter	Type	Description
app_name	str	Name of the Connector Integration

Returns

• list[BaseTool]: List of generated tools

Raises

• ValueError: If app_name is not in the supported modules list

Known Limitations

Some tools generated by the Tools Generator may have limitations when dealing with file operations, particularly binary data handling. This is common with connectors that involve file uploads, downloads, or binary data processing.

Example

For support and questions for example apps you can visit our hello world example apps

Support

For support and questions:

• Check the GL Connectors API documentation

• Contact the GL Connectors Team