Help Center

Appearance

Connect AI Chatbots (ChatGPT, Claude, Gemini, Spark) ​

Pagge provides seamless integration with leading AI Chatbots and Conversational Assistants—including OpenAI ChatGPT, Anthropic Claude, Google Gemini, and Spark / Custom AI Agents.

With AI Chatbots connected, store owners can inspect analytics, manage products and inventory, update themes, adjust layouts, look up orders, and control store operations using plain text instructions in their favorite AI interface.

Key Capabilities ​

When connected to Pagge, your AI Chatbot can help you:

  • Product & Inventory Management: Create new products, upload images, set variants, update prices, and track stock levels.
  • Store Customization & Design: Select themes, configure homepage layouts, edit hero banners, update announcement bars, and adjust color palettes.
  • Analytics & Insights: Ask real-time questions about daily sales, conversion rates, traffic sources, top-selling products, and active orders.
  • Order Operations: Look up customer orders, check fulfillment status, and retrieve customer shipping details.
  • Content & Pages: Write and publish custom pages, store policies (Privacy, Terms, Refund), blog posts, and FAQs.

1. ChatGPT & Custom GPTs ​

You can connect Pagge directly to OpenAI ChatGPT or build a Custom GPT tailored to your brand.

Connecting via Remote MCP Server ​

  1. In ChatGPT, navigate to Custom GPTs > Create a GPT (or edit an existing GPT).
  2. Under Configure > Actions, add a new Action.
  3. Set the Schema / Server URL to Pagge's Remote MCP endpoint:
    http
    https://pagge.co/mcp
  4. Under Authentication, select OAuth 2.0 (or API Key / Bearer Token).
  5. Obtain your API key from your Pagge dashboard under Account & Settings > API Tokens.
  6. Save and test the action. You can now prompt ChatGPT:

    "Show me my top 5 selling products today on Pagge.""Add a new product named 'Organic Coffee Beans' priced at $19.99 with stock set to 50."

2. Anthropic Claude ​

Connect Claude (Claude Desktop or Claude Web) to manage your Pagge store using Model Context Protocol (MCP).

Configuration Setup ​

Add the Pagge MCP server to your claude_desktop_config.json:

json
{
  "mcpServers": {
    "pagge": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-fetch",
        "https://pagge.co/mcp"
      ],
      "env": {
        "PAGGE_API_TOKEN": "YOUR_PAGGE_API_TOKEN"
      }
    }
  }
}

Once connected, ask Claude:

"Audit my store design on Pagge and suggest high-converting layout tweaks.""Update my hero banner subtitle to 'Free Shipping on all orders over $50'."

3. Google Gemini ​

Connect Google Gemini via Custom Extensions or API Function Calling to interact with Pagge.

Setup Instructions ​

  1. In your Gemini developer workspace or API Integration setup, add Pagge's REST API specification:
    http
    https://pagge.co/api/v1/openapi.json
  2. Set authentication header:
    http
    Authorization: Bearer YOUR_PAGGE_API_TOKEN
  3. Enable function calling for Store Management, Products, and Analytics endpoints.

4. Spark & Custom AI Agents ​

For developers building autonomous agents with Spark AI, Cursor, LangChain, or AutoGPT:

  • Protocol: Standard MCP (Model Context Protocol 2026-07-28) or OpenAPI REST endpoints.
  • Endpoint: https://pagge.co/mcp or https://pagge.co/api/v1
  • Authentication: OAuth 2.0 PKCE flow or Bearer API tokens.

Example Agent Instructions ​

"Check Pagge orders every morning at 9 AM, summarize pending shipments, and notify me on Telegram."

Security & Access Control ​

  • Granular Permissions: AI chatbots only access store data authorized by your logged-in Pagge account.
  • Safe State Changes: High-impact actions (such as bulk deletions or account settings changes) require explicit confirmation before execution.
  • Revoking Access: You can revoke any API token or OAuth grant instantly from your Pagge Account & Settings page.