MCP 服务器

@gridinsoft/mcp-inspector

MCP server that enables AI agents to verify website safety, detect phishing, malware, and assess domain reputation using GridinSoft's threat intelligence.

README

@gridinsoft/mcp-inspector

MCP (Model Context Protocol) server for domain and URL security analysis powered by GridinSoft Inspector.

🛡️ Trust & Safety for AI

This MCP server allows your AI agent (Claude, Cursor, etc.) to verify any website or link. It helps answer the critical question: "Can I trust this site?"

✨ Key Features

Domain Reputation - Check if a site is well-known, safe, or suspicious.
Phishing Detection - Identify fraudulent pages designed to steal credentials.
Malware Scanning - Detect links hosting harmful files or scripts.
Real-time Engine - Access GridinSoft's global threat database.

🎁 Free Plan

Every new account gets 100 API credits per month for free. No credit card required.

🚀 Real-World Scenarios

Integrate GridinSoft's threat intelligence directly into your decision-making process:

Secure Development: Automatically audit third-party API endpoints, documentation links, and CDNs before adding them to your codebase.
Automated Documentation Review: Ask your AI to scan all external links in a README or technical specification to ensure no broken or malicious redirects exist.
Incident Response: Paste suspicious logs or email headers into your IDE and have the AI immediately categorize mentioned domains (e.g., C2 servers, phishing drops).
Proactive Security Layers: Instruct your AI agents to "never summarize content from domains with a reputation score below 70" to prevent accidental exposure to malicious content.
Competitive Intelligence: Safely explore new industry sites and verify the legitimacy of emerging platforms mentioned in market research tasks.
E-Commerce Safety: Before your AI agent recommends a product or uses prices from an unfamiliar online store, have it verify the shop's trust score first. Avoid scam sites that offer "too good to be true" deals.

🛠️ Connection Guide

1. Get your API Key

Sign up at inspector.gridinsoft.com.
Grab your key from the Profile page.

2. Connect to your AI

Claude Desktop

Add to your config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "gridinsoft": {
      "command": "npx",
      "args": ["-y", "@gridinsoft/mcp-inspector"],
      "env": {
        "GRIDINSOFT_API_KEY": "your_api_key_here"
      }
    }
  }
}

Cursor or Windsurf IDE

Open Settings -> Models -> MCP.
Add a new command: npx -y @gridinsoft/mcp-inspector.
Set environment variable GRIDINSOFT_API_KEY.

🤖 Example Prompts

Try asking your AI these questions:

"Is the domain example-crypto-site.com safe to use?"
"I received a link to http://bit.ly/xxxx. Scan it for phishing before I open it."
"What is the reputation of gridinsoft.com?"
"Should I trust this website? It looks like a login page but the URL is strange."

🔧 Available Tools

`inspect_domain`

Comprehensive domain security profile. Returns reputation (0-100), vendor verdicts, hosting/IP info, registrar history, and detailed security tags.

Web Version: inspector.gridinsoft.com/domain-checker
Input: { "domain": "example.com" }
Cost: 3 credits

`scan_url`

Deep forensic URL analysis. Extracts metadata, effective URL, full redirect chain, DNS/SSL details, and forensic data (screenshots, network traffic, cookies, console logs for paid plans).

Web Version: inspector.gridinsoft.com/url-scanner
Input: { "url": "https://example.com/login" }
Cost: 10 credits