xpander.ai SDK Ecosystem

The xpander.ai ecosystem offers several packages to help you build, manage, and integrate with AI agents. Each package serves a specific purpose in the xpander platform.

​

Available Packages

​

Installation Guides

​

xpander-sdk

Available in multiple languages:

pip install xpander-sdk

from xpander_sdk import XpanderClient

# Initialize with API key
client = XpanderClient(api_key="your-xpander-api-key")

pip install xpander-sdk

from xpander_sdk import XpanderClient

# Initialize with API key
client = XpanderClient(api_key="your-xpander-api-key")

npm install @xpander-ai/sdk
# or
yarn add @xpander-ai/sdk

import { XpanderClient } from '@xpander-ai/sdk';

// Initialize with API key
const client = new XpanderClient({ apiKey: "your-xpander-api-key" });

using Xpander.SDK;

// Initialize with API key
var client = new XpanderClient("your-xpander-api-key");

import ai.xpander.sdk.XpanderClient;

// Initialize with API key
XpanderClient client = new XpanderClient("your-xpander-api-key");

The SDK provides advanced features including:

• AGS (Function Calling Dependencies): Manage complex function calling dependencies
• Scheme Editor: Visual editor for function calling overides (Input Scheme and Output Scheme)
• Managed Human Interfaces: Integrations with Slack and other collaboration platforms
• Cross-Framework Compatibility: Work seamlessly with various LLM providers and Agent Frameworks

​

xpander-cli

Available in multiple languages:

# Install the CLI
pip install xpander-cli

# Login to xpander
xpander login

# List available agents
xpander agents list

# Create a new agent
xpander agents create --name "My Agent" --type regular

# Install the CLI
pip install xpander-cli

# Login to xpander
xpander login

# List available agents
xpander agents list

# Create a new agent
xpander agents create --name "My Agent" --type regular

# Install the CLI
npm install -g @xpander-ai/cli
# or
yarn global add @xpander-ai/cli

# Login to xpander
xpander login

# List available agents
xpander agents list

# Create a new agent
xpander agents create --name "My Agent" --type regular

# After installation
xpander login
xpander agents list
xpander agents create --name "My Agent" --type regular

# After installation
xpander login
xpander agents list
xpander agents create --name "My Agent" --type regular

The CLI allows you to manage your xpander.ai resources directly from the terminal.

​

xpander-mcp-remote

A command-line utility primarily used with npx to connect MCP clients like Claude Desktop to xpander.ai services.

# Install globally (optional, as it's typically used with npx)
npm install -g xpander-mcp-remote

Typical usage in Claude Desktop configuration:

{
  "mcpServers": {
    "xpander": {
      "command": "npx",
      "args": [
        "xpander-mcp-remote",
        "https://mcp.xpander.ai/your-mcp-url/"
      ]
    }
  }
}

The xpander-mcp-remote package creates a bridge between MCP clients and the xpander.ai platform, enabling:

• MCP Composition: Selectively expose operations from multiple connectors
• Agent Access: Trigger AI agents from MCP clients
• Connector Integration: Connect MCP clients to specific connectors like GitHub, Slack, etc.

For detailed setup instructions, refer to the MCP Integration documentation.

​

xpander-utils (Python only)

pip install xpander-utils

# Use with various frameworks
import xpander_utils

# Example with LangGraph
from xpander_utils.integrations import langgraph_connector
graph = langgraph_connector.create_xpander_graph("my-agent-id")

Utilities library for integrating xpander with AI agent frameworks such as:

• Hugging Face
• CrewAI
• ChainLit
• LangGraph
• And more

​

Core Documentation

​

Additional Resources