ai-testing-mcp

概述

AI Testing MCP 是一个专业的模型上下文协议(MCP)服务器，为AI/ML系统提供标准化的测试方法、评估指标和自动化测试工作流。它实现了MCP规范，可与AI开发工具无缝集成，提供全面的测试功能，包括单元测试、集成测试、性能测试、安全测试和质量测试，并包含准确性、质量、安全性、性能和成本效率等详细评估指标。

试试问 AI

装完之后，这里有 3 个你可以让 AI 做的事：

• run_test_suiterun_test_suite(model, testCategory, testCases)

  Execute a comprehensive test suite for an AI model

  试试:Run accuracy tests on GPT-4 with these test cases试试:Test the performance of Claude-3 with the provided scenarios
• evaluate_outputevaluate_output(output, expected, metrics)

  Evaluate AI model outputs against metrics

  试试:Evaluate this model response for coherence and relevance试试:Check if the output matches the expected answer with exact match
• generate_test_casesgenerate_test_cases(scenario, count)

  Generate test cases for specific scenarios

  试试:Generate 10 test cases for prompt injection security testing试试:Create 5 test cases for mathematical accuracy evaluation

可对比工具

安装

# 克隆仓库
git clone https://github.com/groovy-web/ai-testing-mcp.git
cd ai-testing-mcp

# 安装依赖
npm install

# 配置环境
cp .env.example .env
# 编辑您的API密钥

# 启动MCP服务器
npm start

{
  "mcpServers": {
    "ai-testing": {
      "command": "node",
      "args": ["/path/to/ai-testing-mcp/dist/index.js"],
      "env": {
        "OPENAI_API_KEY": "your-key",
        "ANTHROPIC_API_KEY": "your-key"
      }
    }
  }
}

ai-testing-mcp 对比