purpleroc/mcp-security-inspector
Developer toolsThis page was translated automatically. Technical details may differ from the original. View the original
A Chrome extension tool for detecting the security of Model Context Protocol (MCP).
Overview
MCP Security Inspector is a Chrome extension designed for remote servers, offering two main capabilities: - : Connect to SSE / Streamable HTTP servers to browse and recall tools, resources, and prompts, supporting custom headers and multiple authentication methods - : Combine active scanning and passive monitoring, integrate LLMs such as OpenAI, Claude, Gemini, and others, automatically generate test cases and output risk reports Configurations can be configured between tools such as Cursor, Claude Desktop, VS Code, and this extension, facilitating the reuse of existing MCP configurations for security auditing. Detailed introduction: "MCP Security Inspector: AI-Driven MCP Protocol Security Detection Tool" - 🔌 : SSE / Streamable HTTP； Invoke tools, read resources, get prompts - 📥 Standard mcp.json format, compatible with Cursor, Claude Desktop, VS Code, Windsurf; Automatic filtering of local stdio processes - 🧠 Integrated with OpenAI, Claude, Gemini, and other LLM services - 🔄 : Active scanning + passive monitoring, comprehensive coverage of MCP interactions - 🎯 : Unified 📋 detection framework for tools, prompts, and resources - : {{auto}} automatic beautification, dual columns for raw and formatted materials, wildcard path templates - 🌐 : Full bilingual interface in Chinese and English - 📊 : Real-time generation of detailed security analysis reports - 🛡️ : Local detection, intelligent blocking of sensitive information Using a unified security detection engine, three different types of MCP components—tools, tips, and resources—are incorporated into a unified security detection framework: - User-proactive comprehensive security scans - LLM-enhanced in-depth security analysis - intelligent test case generation and execution - detailed security reports and remediation recommendations - Real-time monitoring of MCP calls and interactions - Automatically detecting potential security threats - Real-time security alerts and logging - History and trend analysis
Overview (machine translated)
AI-enhanced security detection Chrome extension designed specifically for Model Context Protocol (MCP) servers. [Project completed by Cursor]
English | Chinese
🚀 Project Overview
MCP Security Inspector is a Chrome extension designed for Model Context Protocol (MCP) remote servers, offering two main capabilities: protocol debugging and AI-enhanced security detection:
- Protocol Debugging: Connect to an SSE / Streamable HTTP server to browse and call tools, resources, and prompts, supporting custom headers and multiple authentication methods
- Security Detection: Combines active scanning and passive monitoring, integrating LLMs such as OpenAI, Claude, Gemini, to automatically generate test cases and output risk reports
Configurations can be imported/exported standards mcp.json between tools such as Cursor, Claude Desktop, VS Code, and this extension, making it easy to reuse existing MCP configurations for security auditing.
Detailed introduction: “MCP Security Inspector: AI-Driven MCP Protocol Security Detection Tool”
✨ Core Features
- 🔌 MCP Protocol Browser: SSE / Streamable HTTP; Call tools, read resources, obtain prompts
- 📥 Configuration Import and Export: Standard
mcp.jsonformat, compatible with Cursor, Claude Desktop, VS Code, Windsurf; Automatically filters local stdio processes - 🧠 AI-Enhanced Detection: Integrates multiple LLM services such as OpenAI, Claude, Gemini, and more
- 🔄 Dual Detection Mode: Active scanning + passive monitoring, fully covering MCP interaction
- 🎯 Unified Detection Engine: A unified detection framework for three types of MCP components: tools, prompts, and resources
- 📋 Configurable Result Display:
{{auto}}Auto beautification, dual columns for raw and formatted formats, wildcard path templates - 🌐 Multilingual Support: Full bilingual interface in Chinese and English
- 📊 Intelligent Reports: Generate detailed safety analysis reports in real time
- 🛡️ Privacy Protection: Local detection to intelligently shield sensitive information
🏗️ Technical Architecture
Unified Security Detection Engine
Using a unified security detection engine, three different types of MCP components—tools, tips, and resources—are incorporated into a unified security detection framework:
Dual Detection Mode
🔍 Active scanning mode
- Comprehensive user-initiated security scans
- LLM-enhanced in-depth security analysis
- Intelligent test case generation and execution
- Detailed safety reports and repair recommendations
📡 Passive Monitoring Mode
- Real-time monitoring of MCP calls and interactions
- Automatically detects potential security threats
- Real-time security alerts and logging
- Historical records and trend analysis
Testing Process
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 →