Agora

Agora MCP connects your AI assistants to SearchAgora to search, compare, and purchase products across the web. It enables natural conversations with your AI to discover items, manage a shopping experience, and prepare purchases directly through your assistant.

How to use

You interact with Agora MCP through your MCP client (Cursor or Claude Desktop). After configuring the MCP servers, you can ask your AI to search for products, refine results, and eventually add items to a cart or checkout. Start with simple search queries like “find a cool t-shirt” and progressively use advanced filters such as price ranges, pagination, and sorting to tailor results.

To refine results, instruct the AI with more specific prompts such as “Show me headphones under $100 sorted by highest rating.” The system supports price filtering, page navigation, and custom sorting to help you discover the best options.

As new features roll out, you will be able to add items to your cart, view the cart contents, and complete purchases directly from the AI interface. Follow the prompts from your AI to perform these actions once they become available.

How to install

Prerequisites: ensure you have an MCP client like Cursor or Claude Desktop installed, UV available, and a payment method through an L402-compatible client (such as Fewsats). Then follow the steps below to configure access to Agora MCP.

1. Install the MCP client (Cursor or Claude Desktop) on your computer if you haven’t already.

2. Set up MCP server configurations for your chosen client. Use the provided configuration blocks to register Agora MCP and the Fewsats integration as MCP servers.

3. Ensure you have a FEWSATS_API_KEY if you plan to enable the Fewsats MCP integration. Keep this key secure and do not expose it in shared environments.

Configuration blocks you will use

{
  "mcpServers": {
    "AgoraCursor": {
      "command": "uvx",
      "args": [
        "agora-mcp"
      ]
    },
    "FewsatsCursor": {
      "command": "env",
      "args": [
        "FEWSATS_API_KEY=YOUR_FEWSATS_API_KEY",
        "uvx",
        "fewsats-mcp"
      ]
    }
  }
}

Additional sections

The following MCP server configurations are provided for different clients and development scenarios. Each entry is a complete, actionable setup you can paste into your client’s configuration area.

MCP server configurations you will use

Agora MCP supports multiple connection methods depending on your client. Use the following stdio configurations to run the MCP locally via your preferred client. Replace YOUR_FEWSATS_API_KEY with your actual FEWSATS API key when using the Fewsats integration.

{
  "AgoraCursor": {
    "type": "stdio",
    "name": "agora_cursor",
    "command": "uvx",
    "args": ["agora-mcp"]
  },
  "FewsatsCursor": {
    "type": "stdio",
    "name": "fewsats_cursor",
    "command": "env",
    "args": [
      "FEWSATS_API_KEY=YOUR_FEWSATS_API_KEY",
      "uvx",
      "fewsats-mcp"
    ]
  },
  "AgoraClaude": {
    "type": "stdio",
    "name": "agora_claude",
    "command": "uvx",
    "args": ["agora-mcp"]
  },
  "FewsatsClaude": {
    "type": "stdio",
    "name": "fewsats_claude",
    "command": "env",
    "args": [
      "FEWSATS_API_KEY=YOUR_FEWSATS_API_KEY",
      "uvx",
      "fewsats-mcp"
    ]
  },
  "AgoraDev": {
    "type": "stdio",
    "name": "agora_dev",
    "command": "uv",
    "args": [
      "--directory",
      "/path/to/your/agora-mcp",
      "run",
      "agora-mcp"
    ]
  }
}

Security and best practices

Keep your FEWSATS_API_KEY secret and never expose it in client-side configurations or shared environments. Use environment-variable-based configurations whenever possible and apply principle of least privilege to your API keys.

Notes and troubleshooting

If you don’t see results, verify that the MCP server processes are running and accessible from your MCP client. Check that the uvx/uv commands are available in your environment and that the agorа-mcp and fеwsats-mcp services are reachable. Ensure your API key is correctly entered where required.

Available tools

search

Performs a product search with parameters such as query, results per page, page, price range, and sorting.

refine

Applies advanced search filters like price, sorting order, and pagination to refine results.

addToCart

Adds a selected product to the shopping cart (coming soon in the MCP).

viewCart

Displays the current shopping cart contents (coming soon in the MCP).

checkout

Initiates the checkout process to complete a purchase (coming soon in the MCP).

saveFavorites

Saves favorite products for quick access (coming soon in the MCP).

trackOrder

Tracks order status after purchase (coming soon in the MCP).