groovy-web/ai-testing-mcp

View on GitHub ↗

MCP server for comprehensive AI testing, evaluation, and quality assurance

19 ★1 forksUpdated 5mo ago

What you need to know

MCP server providing standardized testing, evaluation metrics, and QA workflows for AI/ML systems, including unit, integration, performance, security, and quality tests.

Install

npm install
npm start

Usage

  • Add the server to your MCP client and use run_test_suite, evaluate_output, and generate_test_cases

Key features

  • Unit, integration, performance, security, and quality tests
  • Evaluation metrics: accuracy, quality, safety, performance, cost
  • Tools: run_test_suite, evaluate_output, generate_test_cases
  • MCP standard protocol integration
  • Requires OpenAI and/or Anthropic API keys

Best for

Testing and evaluating AI models via MCP

Caveats

  • Requires API keys in .env
Platforms: Node.jsClients: MCP-compatible clients
Documentation ↗

Reviewed 2026-08-11

Topics

ai-testingautomationevaluationmcpmodel-context-protocolquality-assurancetestingtypescript
Stars
19★
Forks
1
Language
License
MIT
Created
2026-03-24
Last push
2026-03-24