dest1ny-sec/luvv-mcp

This page was translated automatically. Technical details may differ from the original. View the original

42 stars 2 forks Quality 81 Trend 20MIT

Luvv-MCP is an AI-native front-end security analysis engine. After integrating Claude Code, you can simply say, "Help me audit this website," and AI can automatically identify the technology stack (covering 18,000 rules for CMS/framework/CDN/middleware), leak scanning (Alibaba Cloud AK, GitHub Token, Webhook, JWT, and 60+ types of sensitive information), vulnerability audits (XSS, open redirect, CSRF, IDOR, hard code passwords, 6...

Overview

Traditional front-end security tools are GUI: Burp / Wappalyzer switching back and forth. Luvv-MCP lets AI call directly. You no longer need to switch GUIs; just tell AI "Look what example.com used." - Sniffing rules + fingerprint rules + leak detection - Access Claude Code / Cursor / any client supporting MCP, - No need to switch to GUIs like Burp / Wappalyzer — directly tell the AI, "Look what example.com is using" It is a front-end security analysis engine for AI agents, integrating five capabilities—site sniffing, leak scanning, vulnerability auditing, fingerprint identification, and asset collection—into the Model Context Protocol tool. Once you connect Claude Code / Cursor / any MCP-supporting client, you no longer need to switch to GUIs like Burp or Wappalyzer—just say to the AI: Tools What to Do Number of Rules When to Debug luvv_site_sniff Tech Stack Identification: CMS/Framework/JS Library/CDN/Security Head 7,183 Target Identified First luvv_leak_scan Leak Scan: AK/SK/Token/Webhook/JWT/PII 60+ After obtaining JS source code luvv_vuln_audit Vulnerability Audit: XSS/Redirect/CSRF/IDOR/Hard Coding 6 Classes After all prerequisites are complete luvv_fingerprint_scan Fingerprinting: Server/Middleware/OA/Network Device 11,596 When device model / Known vulnerability association is required luvv_collect_assets Code Collection: JS + SourceMap automatic restore — when conducting deep audits :site_sniff → collect_assets → vuln_audit → leak_scan — AI knows this trick, so you can change it yourself. Restart Claude Code. The MCP panel will display luvv-mcp · connected · 5 tools。 As long as the stdio MCP protocol is supported, configure according to the respective documentation. Luvv-MCP uses the official @modelcontextprotocol/SDK and offers the broadest compatibility.

Overview (machine translated)

⚠️ > For security research use only · Sniffing/auditing behaviors must be performed in an authorized environment · Unauthorized testing is strictly prohibited

It is not a Burp replacement, but rather wrapping front-end security capabilities into an MCP tool

Traditional front-end security tools are GUI: Burp / Wappalyzer switching back and forth. Luvv-MCP packages 5 core capabilities (sniffing/fingerprinting/leak/vulnerability auditing/asset acquisition) into an MCP tool, allowing AI to call directly. You no longer need to switch GUIs and just tell AI, “Look what example.com used.”

7,183 Sniff rule + 11,596 Fingerprint rule + 60+ Leak detection
Access Claude Code / Cursor / Any client that supports MCP, AI direct call
No need to switch to GUIs like Burp or Wappalyzer — just tell the AI, “Look what example.com used.”

What is it?

Luvv-MCP is a front-end security analysis engine for AI agents, integrating five capabilities—site sniffing, leak scanning, vulnerability auditing, fingerprinting, and asset collection—into the Model Context Protocol tool.

Once you connect Claude Code / Cursor / any MCP-supporting client, you no longer need to switch to GUIs like Burp or Wappalyzer—just say to the AI:

AI will select tools, build parameters, and summarize reports on its own. 5 tools cover the complete chain of front-end security "identification→ collection→ auditing, → reporting."

Compared to traditional front-end scanners

Dimension	Traditional GUI Tools	Luvv-MCP
Interaction Methods	Click / Configuration Panel	Natural Language
Data Stream	Output PDF / HTML	Structured JSON, AI can be processed in secondary processing
Orchestration	Manual Serial Connection	AI Automatic Selection Tool + Parameter Selection
Integration	Standalone software	Embedded Claude Code / Cursor Any IDE
Regular quantity	Built-in standalone	7,183 sniffing + 11,596 fingerprints + 60+ leaks

✨ Five MCP tools

Tools What are you doing? Rule number When will the transfer be made?

luvv_site_sniff Tech stack identification: CMS/framework/JS library/CDN/security header 7,183 Adjust your target first

luvv_leak_scan Leak scanning: AK/SK/Token/Webhook/JWT/PII 60+ After obtaining the JS source code,

luvv_vuln_audit Vulnerability audit: XSS/redirect/CSRF/IDOR/hard code Category 6

This page was translated automatically. Technical details may differ from the original.

Install

This server does not publish a one-line install command.

Open the repository installation guide New to MCP? Start with the mcpskill.org guide →