Browse MCP servers — page 91
Explore the next set of open-source servers for connecting AI tools to data and workflows.
MCPAuth
oidebrett
MCPAuth is an OAuth2.1-based authentication and authorization gateway for secure enterprise Model Context Protocol (MCP) integrations. It supports external identity providers like Google OAuth and Keycloak. Configuration requires environment variables for OAuth client credentials and Keycloak settings.
Metabase MCP 服务器
hyeongjun-dev
一个 Model Context Protocol (MCP) 服务器,集成了 AI 助手与 Metabase 分析平台。配置需要环境变量,包括 METABASE_URL 以及 METABASE_API_KEY 或 METABASE_USER_EMAIL 和 METABASE_PASSWORD 以进行身份验证。
Mobile automation iOS MCP server
iHackSubhodip
基于FastMCP 2.0和清晰架构的现代iOS自动化服务器。需要macOS、Python 3.11+、配备iOS模拟器的Xcode,以及用于Appium的Node.js。支持本地开发和通过Railway进行云部署。
Modes MCP Server
mkc909
用于管理Roo的自定义操作模式的MCP(模型上下文协议)服务器,提供对模式配置和管理的编程控制。需要通过MODES_CONFIG_PATH环境变量设置自定义模式配置JSON文件的路径。
Qiniu MCP Server
qiniu
A MCP server project of Qiniu. Supports accessing Qiniu Cloud storage, intelligent multimedia services, and CDN operations via Model Context Protocol. Requires external configuration via environment variables for Qiniu credentials and bucket info.
ramp-mcp: 一个 Ramp MCP 服务器
ramp
用于通过开发者 API 获取和分析数据或执行任务的 Ramp Model Context Protocol 服务器。支持使用临时内存中的 sqlite 数据库的 ETL 管道。通过配置环境变量 RAMP_CLIENT_ID、RAMP_CLIENT_SECRET 和 RAMP_ENV 来控制行为。
Redis MCP 服务器
gongrzhe
用于与 Redis 数据库交互的 Redis 模型上下文协议(MCP)服务器实现。通过标准化工具,使大型语言模型(LLMs)能够与 Redis 键值存储进行交互。
revula
president-xd
Production-grade MCP server for universal reverse engineering automation. Connect Claude Desktop, MCP-compatible IDEs, or custom tooling to a complete reverse engineering backend. One server, every RE tool, orchestrated through the Model Context Protocol.
Roam MCP
Roam-Research
A Model Context Protocol (MCP) server and CLI for Roam Research. Requires Roam Research desktop app (local API not available in web version).
Sugar
roboticforce
Sugar is an autonomous AI development tool for Claude Code that continuously executes tasks such as building features, fixing bugs, running tests, and integrating with GitHub. It supports task queue management, autonomous execution, and smart discovery. Installable via PyPI package 'sugarai'.
The Training Platform for Specialized Model
overmind-core
Overmind trains models you own, on data only you have. It provides a context graph, observability, data workshop, evaluations, optimizer, training, and inference capabilities. The server can be self-hosted with Docker Compose and requires configuration via environment variables documented in .env.example.