Version: 2608.1

Sparrow On-Demand MCP

MCP (Model Context Protocol)‌ is a standard protocol that enables safe interaction with a variety of API services through an AI LLM (Large Language Model). Through this technology, Sparrow On-Demand MCP makes it easy to use the analysis features provided by Sparrow On-Demand from the AI plugin in your IDE.

Sparrow On-Demand MCP provides the following features.

Security analysis of a single file or a folder
Checking analysis progress
Looking up analysis results and analysis rules

MCP architecture

Execution flow

note

It is a good idea to run a test analysis with a sample project. Use the test analysis to verify the authentication, analysis request, and response flows in advance.

👤 CLIENTClient

The MCP client that connects your local IDE to the Sparrow On-Demand MCP server.
You can invoke it by specifying Sparrow On-Demand MCP when you enter a command in the LLM chat panel of your local IDE.

📦 MCP serverMCP server

The NPM-based MCP server provided by Sparrow On-Demand.
It can relay requests and responses between the LLM chat panel of the local IDE, the LLM server, and the Sparrow On-Demand server.

🧠 ONDEMANDOn-Demand

The Sparrow On-Demand server that performs the requested analysis.
The Sparrow On-Demand server runs on AWS and performs source code analysis, open source analysis, and web vulnerability analysis.

🧾 LLM serverLLM server

An LLM (Large Language Model) server that you install locally or that is provided as a commercial cloud service.
It generates secure code based on the Sparrow On-Demand analysis results.

Installation

Prerequisites

Install Node.js 22 or later and npm
A Sparrow On-Demand API token is required

Tip: See Token issuance.
Connect your IDE to an LLM, for example through a plugin

If you already use another IDE and LLM, there are no additional requirements. This section explains how to connect Sparrow On-Demand MCP in Cursor IDE to a local Ollama instance. For how to install and configure Ollama, see the Ollama official documentation.

Installing the MCP package

Install the MCP package with npm. You can install it globally or locally.

# Global installation
npm install -g @sparrowai/sparrow-mcp

# Local installation
npm install @sparrowai/sparrow-mcp

Creating the MCP configuration file

Create the MCP configuration file (mcp.json) in your IDE and add the configuration including your Sparrow On-Demand API token.

//Global installation
{
  "mcpServers": {
    "sparrow-mcp": {
      "command": "node",
      "args": ["C:\\Users\\<username>\\AppData\\Roaming\\npm\\node_modules\\@sparrowai\\sparrow-mcp\\dist\\src\\server.js"],
      "env": {
        "SPARROW_API_KEY": "your-ondemand-token",
        "SPARROW_API_URL": "https://ondemand.sparrowcloud.ai",
        "OLLAMA_BASE_URL": "http://localhost:11434",
        "OLLAMA_MODEL": "gpt-oss:20b",
        "NODE_ENV": "development"
      }
    }
  }
}

// Local installation
{
  "mcpServers": {
    "sparrow-mcp": {
      "command": "node",
      "args": ["./node_modules/@sparrowai/sparrow-mcp/dist/src/server.js"],
      "env": {
        "SPARROW_API_KEY": "API_KEY",
        "SPARROW_API_URL": "https://ondemand.sparrowcloud.ai",
        "OLLAMA_BASE_URL": "YOUR_OLLAMA_SERVER_URL",
        "OLLAMA_MODEL": "gpt-oss:20b",
        "NODE_ENV": "development"
      }
    }
  }
}

//Global installation
{
  "mcpServers": {
    "sparrow-mcp": {
      "command": "node",
      "args": ["C:\\Users\\<username>\\AppData\\Roaming\\npm\\node_modules\\@sparrowai\\sparrow-mcp\\dist\\src\\server.js"],
      "env": {
        "SPARROW_API_KEY": "your-ondemand-token",
        "SPARROW_API_URL": "https://ondemand.sparrowcloud.ai",
        "OLLAMA_BASE_URL": "http://localhost:11434",
        "OLLAMA_MODEL": "gpt-oss:20b",
        "NODE_ENV": "development"
      }
    }
  }
}
 
// Local installation
{
  "mcpServers": {
    "sparrow-mcp": {
      "command": "node",
      "args": [".\\node_modules\\@sparrowai\\sparrow-mcp\\dist\\src\\server.js"],
      "env": {
        "SPARROW_API_KEY": "API_KEY",
        "SPARROW_API_URL": "https://ondemand.sparrowcloud.ai",
        "OLLAMA_BASE_URL": "YOUR_OLLAMA_SERVER_URL",
        "OLLAMA_MODEL": "gpt-oss:20b",
        "NODE_ENV": "development"
      }
    }
  }
}

Now save the mcp.json file and restart Cursor IDE.

Usage