1. Conduid
  2. AI
  3. dealx/mcp-server
MCP server · AI

dealx/mcp-server

MCP Server for DealX platform

Unclaimed MIT last commit a year ago ai
52Fair

Scored 6 days ago · breakdown

About dealx/mcp-server

dealx/mcp-server is an MCP server published by DealExpress in the AI category: mCP Server for DealX platform.

Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.

Install

From source
git clone https://github.com/DealExpress/mcp-server

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about dealx/mcp-server

Powered by Claude · Grounded in docs

I know everything about dealx/mcp-server. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

  • ✓README presentThe repository ships documentation.
  • ✓License declaredUsable under a stated license.
  • !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.

3 checks have not been run yet.

README

@dealx/mcp-server

This is a Model Context Protocol (MCP) server for the DealX platform. It allows LLMs to interact with the DealX platform, specifically to search for ads.

Table of Contents

Hosted deployment

A hosted deployment is available on Fronteir AI.

Overview

The DealX MCP Server implements the Model Context Protocol to provide a standardized way for LLMs to interact with the DealX platform. Currently, it supports searching for ads, with plans to add more functionality in the future.

What is MCP?

The Model Context Protocol (MCP) is a standardized way for LLMs to interact with external systems. It provides a structured interface for LLMs to access data and perform actions in the real world. This server implements the MCP specification to allow LLMs to interact with the DealX platform.

Installation

Prerequisites

  • Node.js (v20 or later)
  • npm (v11 or later)

MCP Configuration

To use this server with an LLM like Claude, you need to add it to your LLM's MCP configuration:

  1. Open your LLM's MCP configuration file:

    • Claude Desktop App:
      • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
      • Windows: %APPDATA%\Claude\claude_desktop_config.json
      • Linux: ~/.config/Claude/claude_desktop_config.json
    • Cline (VS Code Extension):
      • ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json
  2. Add the DealX MCP server to the mcpServers section:

    {
      "mcpServers": {
        "dealx": {
          "command": "npx",
          "args": ["-y", "@dealx/mcp-server"],
          "env": {
            "DEALX_API_URL": "https://dealx.com.ua"
          },
          "disabled": false,
          "autoApprove": []
        }
      }
    }
    

Installation via npm

The easiest way to install the DealX MCP Server is via npm:

npm install -g @dealx/mcp-server

Installation for Development

If you want to modify the server or contribute to its development:

  1. Clone the repository:

    git clone <repository-url>
    cd dealx/mcp
    
  2. Install dependencies:

    npm install
    
  3. Create a .env file based on the .env.example file:

    cp .env.example .env
    
  4. Edit the .env file to set the appropriate values:

    # DealX API URL
    DEALX_API_URL=http://localhost:3001
    
    # Optional: Specify the port for the MCP server
    MCP_SERVER_PORT=3100
    
    # Optional: Log level (debug, info, warn, error)
    LOG_LEVEL=info
    
  5. Build the server:

    npm run build
    

Usage

Starting the Server

You can run the server in several ways:

  1. If installed globally:

    node node_modules/@dealx/mcp-server/build/index.js
    
  2. Using npx without installation:

    npx -y @dealx/mcp-server
    
  3. With environment variables:

    DEALX_API_URL=https://dealx.com.ua npx -y @dealx/mcp-server
    
  4. For development:

    npm start
    

Using with an LLM

Once configured in your LLM's MCP settings, you can use natural language to interact with the DealX platform.

Example prompts:

  • "Search for ads on DealX with the query 'laptop'"
  • "Find the newest 5 ads for 'iPhone' on DealX"
  • "Search DealX for apartments in Kyiv"

Available Tools

search_ads

Search for ads on the DealX platform.

Parameters:

  • query (string, optional): Search query string
  • sort (string, optional): Sort order (e.g., "-created" for newest first)
  • offset (number, optional): Pagination offset (starts at 1, default: 1)
  • limit (number, optional): Number of results per page (max 100, default: 30)

Example Usage:

{
  "query": "laptop",
  "sort": "-created",
  "offset": 1,
  "limit": 10
}

Extending the Server

The server is designed to be easily extended with additional tools. Here's how to add a new tool:

  • Define the tool in the TOOLS object in src/index.ts:

    const TOOLS = {
      SEARCH_ADS: "search_ads",
      NEW_TOOL: "new_tool", // Add your new tool here
    };
    
  • Create a new file in the src/tools directory for your tool implementation:

    // src/tools/new-tool.ts
    import { ErrorCode, McpError } from "@modelcontextprotocol/sdk/types.js";
    
    interface NewToolParams {
      // Define your tool parameters here
    }
    
    export async function newTool(params: NewToolParams) {
      try {
        // Implement your tool logic here
    
        return {
          content: [
            {
              type: "text",
              text: JSON.stringify(result, null, 2),
            },
          ],
        };
      } catch (error) {
        // Handle errors
        // ...
      }
    }
    
  • Add the tool to the ListToolsRequestSchema handler in src/index.ts:

    this.server.setRequestHandler(ListToolsRequestSchema, async () => ({
      tools: [
        // Existing tools...
        {
          name: TOOLS.NEW_TOOL,
          description: "Description of your new tool",
          inputSchema: {
            type: "object",
            properties: {
              // Define your tool parameters here
            },
            required: [], // List required parameters
          },
        },
      ],
    }));
    
  • Add the tool to the CallToolRequestSchema handler in src/index.ts:

    this.server.setRequestHandler(CallToolRequestSchema, async (request) => {
      const { name, arguments: args } = request.params;
    
      switch (name) {
        // Existing cases...
        case TOOLS.NEW_TOOL:
          return await newTool(args);
        default:
          throw new McpError(ErrorCode.MethodNotFound, `Unknown tool: ${name}`);
      }
    });
    
  • Import your new tool in src/index.ts:

    import { newTool } from "./tools/new-tool.js";
    

Planned Future Tools

The following tools are planned for future implementation:

  • create_ad: Create a new ad on the DealX platform
  • edit_ad: Edit an existing ad
  • delete_ad: Delete an ad
  • get_threads: Get discussion threads for an ad
  • create_thread: Create a new discussion thread

Development

Project Structure

mcp/
├── build/              # Compiled JavaScript files
├── src/                # TypeScript source files
│   ├── tools/          # Tool implementations
│   │   └── search-ads.ts
│   └── index.ts        # Main server implementation
├── .env                # Environment variables (not in git)
├── .env.example        # Example environment variables
├── package.json        # Project dependencies and scripts
├── tsconfig.json       # TypeScript configuration
└── README.md           # This file

npm Scripts

  • npm run build - Compile TypeScript to JavaScript
  • npm start - Start the server using the compiled JavaScript
  • npm run dev - Start the server in development mode with hot reloading
  • npm run lint - Lint the code using ESLint
  • npm run format - Format the code using Prettier
  • npm test - Run tests

Troubleshooting

Common Issues

Server Not Starting

If the server fails to start, check the following:

  • Make sure you have the correct Node.js version installed
  • Check that all dependencies are installed
  • Verify that the .env file exists and has the correct values
  • Check the console output for error messages

Connection Issues

If the LLM can't connect to the server:

  • Make sure the server is running
  • Check that the MCP configuration in the LLM's settings is correct
  • Verify that the path to the server executable is correct
  • Check that the environment variables are set correctly

API Connection Issues

If the server can't connect to the DealX API:

  • Make sure the DealX API is running
  • Check that the DEALX_API_URL environment variable is set correctly
  • Verify that the API endpoint is accessible from the server

Getting Help

If you encounter issues not covered here, please open an issue against this GitHub repository.

README copy captured when this server was indexed; it may lag the repository. The original is authoritative.

Questions

About dealx/mcp-server

How do I install dealx/mcp-server?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is dealx/mcp-server safe to use with an AI agent?

Its trust score is 52 out of 100 (fair). It passes 2 of the 3 static security checks run so far; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is dealx/mcp-server still maintained?

The last commit was a year ago. That's long enough that you should check whether the maintainer is responding to issues before depending on it.