发现优秀的 MCP 服务器

Tierward

Tierward is a development framework that governs Claude Code sessions through a tiered pipeline with STOP gates, audit skills, and test enforcement, enabling teams to review and approve AI-generated code before merging.

Enhanced Word Document MCP Server

A consolidated MCP server for creating, reading, and manipulating Microsoft Word documents with 25 tools including search/replace, track changes, and document protection.

devbase

Developer Knowledge OS — local-first bimodal workspace for humans (TUI) and AI agents (MCP). Manage Git repos, vault notes, and assets with 19 MCP tools including unified project context queries.

android-dev-mcp

Unified MCP server for Android mobile development — combines project knowledge, Android CLI execution, live dependency versions, and agentic workflow tools for Claude Code.

console-stream-mcp

Enables AI coding agents to access live console logs, errors, and network requests from web applications via a local WebSocket connection, without copying data to chat.

F1 Live MCP server

Provides live Formula 1 data such as driver standings, race results, and schedule for the current season, enabling users to ask about F1 without stale training data.

LibreChat MCP Servers

Okay, here's a translation of instructions for setting up SuperGateway MCP servers in Docker containers for Docker deployments of LibreChat: **Title: 在 Docker 容器中设置 SuperGateway MCP 服务器，用于 LibreChat 的 Docker 部署** **Instructions:** This document outlines the steps to set up SuperGateway MCP (Message Control Plane) servers within Docker containers for use with Docker deployments of LibreChat. This setup allows for scalable and robust message handling within your LibreChat environment. **1. Prerequisites (先决条件):** * **Docker and Docker Compose:** Ensure you have Docker and Docker Compose installed on your system. Refer to the official Docker documentation for installation instructions: [https://docs.docker.com/](https://docs.docker.com/) * **Docker 和 Docker Compose：** 确保您的系统上已安装 Docker 和 Docker Compose。 请参阅官方 Docker 文档以获取安装说明：[https://docs.docker.com/](https://docs.docker.com/) * **LibreChat Docker Deployment:** You should already have a working LibreChat Docker deployment. These instructions assume you are extending an existing setup. * **LibreChat Docker 部署：** 您应该已经有一个可以正常运行的 LibreChat Docker 部署。 这些说明假定您正在扩展现有设置。 **2. Create a Docker Compose File (创建 Docker Compose 文件):** Create a `docker-compose.override.yml` file (or modify your existing `docker-compose.yml` if you prefer) to define the MCP services. This file will define the MCP containers and their configuration. Here's an example: ```yaml version: "3.9" services: mcp1: image: ghcr.io/danny-dann/supergateway:latest # Replace with the correct SuperGateway image container_name: mcp1 ports: - "3001:3000" # Map host port 3001 to container port 3000 environment: - NODE_ENV=production - PORT=3000 - REDIS_HOST=redis # Assuming your Redis service is named 'redis' in your docker-compose file - REDIS_PORT=6379 - MCP_ID=mcp1 # Unique identifier for this MCP instance depends_on: - redis mcp2: image: ghcr.io/danny-dann/supergateway:latest # Replace with the correct SuperGateway image container_name: mcp2 ports: - "3002:3000" # Map host port 3002 to container port 3000 environment: - NODE_ENV=production - PORT=3000 - REDIS_HOST=redis # Assuming your Redis service is named 'redis' in your docker-compose file - REDIS_PORT=6379 - MCP_ID=mcp2 # Unique identifier for this MCP instance depends_on: - redis ``` * **创建 Docker Compose 文件：** 创建一个 `docker-compose.override.yml` 文件（或者如果您愿意，修改您现有的 `docker-compose.yml` 文件）来定义 MCP 服务。 此文件将定义 MCP 容器及其配置。 这是一个例子： ```yaml version: "3.9" services: mcp1: image: ghcr.io/danny-dann/supergateway:latest # 替换为正确的 SuperGateway 镜像 container_name: mcp1 ports: - "3001:3000" # 将主机端口 3001 映射到容器端口 3000 environment: - NODE_ENV=production - PORT=3000 - REDIS_HOST=redis # 假设您的 Redis 服务在您的 docker-compose 文件中命名为“redis” - REDIS_PORT=6379 - MCP_ID=mcp1 # 此 MCP 实例的唯一标识符 depends_on: - redis mcp2: image: ghcr.io/danny-dann/supergateway:latest # 替换为正确的 SuperGateway 镜像 container_name: mcp2 ports: - "3002:3000" # 将主机端口 3002 映射到容器端口 3000 environment: - NODE_ENV=production - PORT=3000 - REDIS_HOST=redis # 假设您的 Redis 服务在您的 docker-compose 文件中命名为“redis” - REDIS_PORT=6379 - MCP_ID=mcp2 # 此 MCP 实例的唯一标识符 depends_on: - redis ``` **Explanation (说明):** * **`version: "3.9"`:** Specifies the Docker Compose file version. * **`version: "3.9"`：** 指定 Docker Compose 文件版本。 * **`services:`:** Defines the services (containers) to be created. * **`services:`：** 定义要创建的服务（容器）。 * **`mcp1` and `mcp2`:** Define two MCP service instances. You can create more instances for increased redundancy and scalability. * **`mcp1` 和 `mcp2`：** 定义两个 MCP 服务实例。 您可以创建更多实例以提高冗余和可扩展性。 * **`image: ghcr.io/danny-dann/supergateway:latest`:** Specifies the Docker image to use for the MCP server. **Important:** Replace this with the correct and up-to-date SuperGateway image name and tag. * **`image: ghcr.io/danny-dann/supergateway:latest`：** 指定用于 MCP 服务器的 Docker 镜像。 **重要提示：** 将其替换为正确且最新的 SuperGateway 镜像名称和标签。 * **`container_name`:** Sets the name of the container. * **`container_name`：** 设置容器的名称。 * **`ports`:** Maps ports from the host machine to the container. Make sure these ports are not already in use. Each MCP instance should have a unique host port. * **`ports`：** 将端口从主机映射到容器。 确保这些端口尚未被使用。 每个 MCP 实例都应该有一个唯一的主机端口。 * **`environment`:** Sets environment variables for the container. * **`environment`：** 设置容器的环境变量。 * **`NODE_ENV=production`:** Sets the Node.js environment to production. * **`NODE_ENV=production`：** 将 Node.js 环境设置为生产环境。 * **`PORT=3000`:** Specifies the port the MCP server listens on within the container. This should match the port exposed by the SuperGateway image. * **`PORT=3000`：** 指定 MCP 服务器在容器中侦听的端口。 这应该与 SuperGateway 镜像公开的端口匹配。 * **`REDIS_HOST=redis`:** Specifies the hostname of the Redis server. This assumes your Redis service is named `redis` in your `docker-compose.yml` file. Adjust this if your Redis service has a different name. * **`REDIS_HOST=redis`：** 指定 Redis 服务器的主机名。 这假设您的 Redis 服务在您的 `docker-compose.yml` 文件中命名为 `redis`。 如果您的 Redis 服务具有不同的名称，请进行调整。 * **`REDIS_PORT=6379`:** Specifies the port of the Redis server. * **`REDIS_PORT=6379`：** 指定 Redis 服务器的端口。 * **`MCP_ID=mcp1` (and `MCP_ID=mcp2`):** A unique identifier for each MCP instance. This is important for the MCP servers to distinguish themselves. * **`MCP_ID=mcp1`（和 `MCP_ID=mcp2`）：** 每个 MCP 实例的唯一标识符。 这对于 MCP 服务器区分自身非常重要。 * **`depends_on: - redis`:** Ensures that the Redis service is started before the MCP services. * **`depends_on: - redis`：** 确保 Redis 服务在 MCP 服务之前启动。 **3. Update LibreChat Configuration (更新 LibreChat 配置):** You need to configure LibreChat to use the MCP servers. This typically involves setting environment variables in your LibreChat Docker container. Consult the LibreChat documentation for the specific environment variables related to SuperGateway/MCP integration. Example variables might include: * `SUPERGATEWAY_ENABLED=true` * `SUPERGATEWAY_URLS=http://localhost:3001,http://localhost:3002` (or the appropriate URLs based on your setup) * **更新 LibreChat 配置：** 您需要配置 LibreChat 以使用 MCP 服务器。 这通常涉及在您的 LibreChat Docker 容器中设置环境变量。 请查阅 LibreChat 文档，了解与 SuperGateway/MCP 集成相关的特定环境变量。 示例变量可能包括： * `SUPERGATEWAY_ENABLED=true` * `SUPERGATEWAY_URLS=http://localhost:3001,http://localhost:3002`（或基于您的设置的相应 URL） **4. Start the Services (启动服务):** Run the following command in the directory containing your `docker-compose.yml` and `docker-compose.override.yml` files: ```bash docker-compose up -d ``` This will start all the services defined in your Docker Compose files, including the MCP servers. * **启动服务：** 在包含您的 `docker-compose.yml` 和 `docker-compose.override.yml` 文件的目录中运行以下命令： ```bash docker-compose up -d ``` 这将启动您的 Docker Compose 文件中定义的所有服务，包括 MCP 服务器。 **5. Verify the Setup (验证设置):** * Check the logs of the MCP containers to ensure they are starting up correctly and connecting to Redis. You can use the following command: ```bash docker logs mcp1 docker logs mcp2 ``` * Verify that LibreChat is able to connect to the MCP servers and that messages are being routed correctly. This may involve testing different message types and observing the behavior of the system. * **验证设置：** * 检查 MCP 容器的日志，以确保它们正确启动并连接到 Redis。 您可以使用以下命令： ```bash docker logs mcp1 docker logs mcp2 ``` * 验证 LibreChat 是否能够连接到 MCP 服务器，以及消息是否正确路由。 这可能涉及测试不同的消息类型并观察系统的行为。 **Important Considerations (重要注意事项):** * **SuperGateway Image:** Always use the latest stable version of the SuperGateway image. Check the SuperGateway repository for updates. * **SuperGateway 镜像：** 始终使用最新稳定版本的 SuperGateway 镜像。 检查 SuperGateway 存储库以获取更新。 * **Networking:** Ensure that the MCP containers and the LibreChat container are on the same Docker network so they can communicate with each other. This is usually handled automatically by Docker Compose. * **网络：** 确保 MCP 容器和 LibreChat 容器位于同一个 Docker 网络上，以便它们可以相互通信。 这通常由 Docker Compose 自动处理。 * **Redis:** The MCP servers rely on Redis for message queuing and coordination. Ensure that your Redis server is properly configured and accessible to the MCP containers. * **Redis：** MCP 服务器依赖 Redis 进行消息排队和协调。 确保您的 Redis 服务器已正确配置并且 MCP 容器可以访问。 * **Scalability:** You can scale the number of MCP instances to handle increased message traffic. Simply add more MCP service definitions to your `docker-compose.yml` file and update the `SUPERGATEWAY_URLS` in your LibreChat configuration accordingly. * **可扩展性：** 您可以扩展 MCP 实例的数量以处理增加的消息流量。 只需将更多 MCP 服务定义添加到您的 `docker-compose.yml` 文件，并相应地更新 LibreChat 配置中的 `SUPERGATEWAY_URLS`。 * **Security:** Consider security implications, especially if exposing the MCP servers to the internet. Use appropriate security measures such as firewalls and authentication. * **安全性：** 考虑安全影响，尤其是在将 MCP 服务器暴露于互联网时。 使用适当的安全措施，例如防火墙和身份验证。 This translation should provide a clear and accurate guide for setting up SuperGateway MCP servers in Docker for LibreChat deployments. Remember to consult the official LibreChat and SuperGateway documentation for the most up-to-date information and specific configuration options. Good luck!

VoyageHacks MCP Server

MCP server for VoyageHacks.com providing travel tools to search articles, get travel gear recommendations, and generate booking links for flights, hotels, eSIMs, VPNs, and credit cards across multiple languages.

yamtrack-mcp

Exposes the Yamtrack REST API as tools for LLMs, enabling media tracking and management through natural language.

agent-news

Enables querying verified AI agent news with citations and confidence scores. Provides tools for search, Q&A, article retrieval, and recommendations about AI agent tools, MCPs, and frameworks.

datadog-mcp

Retrieves logs from Datadog for a specified Kubernetes cluster and namespace, enabling analysis by Claude through natural language queries.

YouTrack MCP Server

Enables AI assistants to interact with JetBrains YouTrack for issue, project, and user management through natural language.

YouTube Transcript MCP (HTTP)

Enables retrieval of YouTube video transcripts/subtitles and titles using yt-dlp, with automatic subtitle conversion to reduce tokens. Supports Streamable HTTP transport for easy integration with frontends like OpenWebUI and Odysseus.

Enterprise Infrastructure & Metrics MCP Server

Provides read-only access to host system metrics (CPU, memory, disk), Docker container health/logs, and sandboxed log file analysis via MCP tools, enabling AI agents to monitor enterprise infrastructure safely.

MCP智能简历投递助手

Enables AI-driven job application automation for LinkedIn and SEEK platforms with intelligent cover letter generation, automated application submission, and application tracking management. Supports anti-detection measures and complies with platform usage policies for safe job hunting automation.

data-quality-cde-mcp-server

Generates and deploys Cloudera Data Engineering (CDE) Airflow jobs from ODCS data contracts, validating live table data in CDW against schema and data-quality rules without WAP staging.

daft-mcp

MCP server that enables AI assistants to search Irish property listings on Daft.ie via the daftlistings library.

axonn-mcp

MCP server for Axonn, enabling access to US energy regulatory filings, real-time ISO prices, and market data.

Gemini Flash MCP - Image Generation for Roo Code

用于 Google Gemini 2.0 Flash 图像生成的 MCP 服务器

pykrx-mcp

Provides Korean stock market data (KOSPI, KOSDAQ, KONEX) including prices, fundamentals, investor trading, short selling, and indices via MCP protocol, enabling natural language queries from AI agents like ChatGPT and Claude.

struct-mcp

Transform data structure definitions into queryable MCP servers, enabling natural language queries about field meanings, data lineage, and structure.

mcp-hubspot

HubSpot MCP server that enables browsing and searching contacts, companies, and deals in your HubSpot workspace.

Remote MCP Server on Cloudflare

A template for deploying authentication-free MCP servers on Cloudflare Workers, enabling custom tool creation and integration with Claude Desktop and AI Playground.

Shiprocket MCP

Enables courier rate checks, order management, shipping, and tracking via Shiprocket's API through natural language.

l6e-mcp

Enables AI coding agents to set budgets per task, check costs before expensive operations, and halt when budget is exhausted, with support for calibration and cloud sync.

Dev Tool MCP

Provides web crawling and browser automation capabilities with support for multiple content formats (HTML, JSON, PDF, screenshots, Markdown), page content extraction, console message monitoring, and network request tracking.

MCP Outlook Server

Enables AI agents to interact with Microsoft Outlook via the Microsoft Graph API for managing emails and calendar events. It allows users to read and send emails, list messages, and create calendar appointments with automatic Teams links.

living-off-the-land-lolbins-mcp-server

Living off the Land intelligence for AI agents. GTFOBins, LOLBAS, LOOBins, LOLDrivers, LOLRMM, WADComs, LOLC2 and more — unified into a single MCP server.

MemoryClaw

Enables AI agents to store and retrieve persistent memories using BM25 search, allowing them to remember past conversations and context across sessions.