Token导航 LogoToken导航TokenDH.com
AeyeWire MCP logo
安全风控未说明官方级别未说明来源级核验

AeyeWire MCP

MCP Server

一个利用qwen/qwen3-coder-30b语言模型进行源代码自动化安全分析的MCP服务,支持多种编程语言并提供IDE集成和命令行工具。

工具数

3

提示词数

0

GitHub Stars

0

资源数

0
静态分析GoAI驱动IDE集成

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

作者 / 组织

ettoremessina

提供方

ettoremessina

最后核验

2026/5/17 20:22

快速接入

先看主来源和安装命令,再打开仓库或文档;下面只保留这个条目的关键接入事实。

详细介绍

AeyeWire MCP Service

A Model Context Protocol (MCP) service that performs automated security analysis on source code using LMStudio with the qwen/qwen3-coder-30b language model.

Features

  • Multi-Language Support: C#, Java, React (TypeScript/JavaScript)
  • MCP Integration: Standard stdio interface for IDE integration
  • CLI Mode: Command-line tool for standalone usage
  • Comprehensive Analysis: 20-25+ security rules per language
  • LLM-Powered: Leverages advanced AI for intelligent security detection

Prerequisites

  • Go 1.21 or higher
  • LMStudio running with qwen/qwen3-coder-30b model
  • LMStudio server accessible at http://localhost:1234 (default)

Installation

  1. Clone the repository:
git clone 
cd aeyewire-mcp
  1. Install dependencies:
make install
  1. Build the binary:
make build

Configuration

Configure via environment variables (all optional with defaults):

export LMSTUDIO_BASE_URL="http://localhost:1234"  # LMStudio server URL
export LMSTUDIO_MODEL="qwen/qwen3-coder-30b"      # Model name
export LMSTUDIO_API_KEY=""                         # API key (if required)
export MCP_SERVER_NAME="aeyewire_mcp"            # Server identifier
export MCP_SERVER_VERSION="1.0.0"                 # Service version

Usage

MCP Server Mode

Run as an MCP stdio server for IDE integration:

./build/aeyewire_mcp

Or with make:

make run

Command-Line Mode

Analyze a specific file:

./build/aeyewire_mcp analyze path/to/file.java

Check service health:

./build/aeyewire_mcp health
# or
make health

List supported languages:

./build/aeyewire_mcp languages
# or
make languages

Show version:

./build/aeyewire_mcp version
# or
make version

MCP Tools

The service exposes three MCP tools:

1. analyze_security

Performs comprehensive security analysis on source code.

Parameters:

  • code (string, required): Source code to analyze
  • file_path (string, optional): File path for context
  • language (string, optional): Language override (csharp, java, react_typescript, react_javascript, auto)

Returns: Markdown-formatted security report

2. health_check

Verifies service health and dependency availability.

Parameters: None

Returns: JSON health status

3. list_supported_languages

Lists all supported programming languages.

Parameters: None

Returns: JSON array of language metadata

Supported Languages

  • C# (.cs) - 20+ security rules
  • Java (.java) - 25+ security rules
  • React TypeScript (.tsx, .ts) - 20+ security rules
  • React JavaScript (.jsx, .js) - 20+ security rules

Development

Format code:

make fmt

Run linter:

make lint

Run tests:

make test

Clean build artifacts:

make clean

Project Structure

.
├── src/
│   ├── AeyeWire_mcp.go           # Main MCP service
│   ├── models/
│   │   └── models.go              # Data models
│   ├── services/
│   │   ├── language_detector.go   # Language detection
│   │   └── llm_service.go         # LLM integration
│   └── analyzers/
│       ├── base_analyzer.go       # Base analyzer
│       ├── java_analyzer.go       # Java analyzer
│       ├── csharp_analyzer.go     # C# analyzer
│       └── react_analyzer.go      # React analyzer
├── docs/
│   └── specifications.md          # Detailed specifications
├── go.mod                         # Go module file
├── Makefile                       # Build automation
└── README.md                      # This file

Security Analysis Coverage

Common Vulnerabilities Detected

  • SQL Injection
  • Command Injection
  • XSS (Cross-Site Scripting)
  • Path Traversal
  • Insecure Deserialization
  • Weak Cryptography
  • Hardcoded Credentials
  • Authentication Bypass
  • CSRF Vulnerabilities
  • And many more...

See docs/specifications.md for complete security rule details.

Troubleshooting

LLM Service Unavailable:

  • Ensure LMStudio is running
  • Verify the model is loaded
  • Check LMSTUDIO_BASE_URL configuration

Language Detection Issues:

  • Provide file_path parameter for extension-based detection
  • Use language parameter to override detection

Analysis Timeout:

  • Large files may take longer to analyze
  • Default timeout is 120 seconds

License

This project is licensed under the MIT License - see the LICENSE file for details.

目录标签

目录标签

静态分析GoAI驱动IDE集成代码安全本地部署多语言支持

接入字段

传输方式(transport,传输协议)

未说明

鉴权方式(authType,认证方式)

api-key

工具数量(toolCount,工具数)

3

资源数量(resourceCount,资源数)

0

提示词数量(promptCount,提示词数)

0

权限和风险

未说明api-key部署方式未说明

接入前请确认传输方式、认证方式和部署位置,并根据实际工具能力限制访问范围。

安装前确认

不要直接授予不必要的文件、网络或账号权限;先核对安装命令和配置内容。

仍需确认:installCommand

来源信息

继续浏览同类 MCP