发现优秀的 MCP 服务器

通过 MCP 服务器扩展您的代理能力,拥有 86,267 个能力。

全部86,267
prompts-mcp-server

prompts-mcp-server

Automatically generates and manages a prompts system for software projects, enabling persistent context for AI coding assistants through project scanning, requirement clarification, and module tracking.

open-ssh-mcp

open-ssh-mcp

An MCP server for managing remote SSH servers, enabling AI agents to execute commands, transfer files, and perform deployment operations securely.

PDF Tools MCP Server

PDF Tools MCP Server

A comprehensive tool server for reading, merging, and extracting content from PDF files via local paths or direct URLs. It enables metadata retrieval, regex searching, and page-specific text extraction with built-in caching and workspace-restricted security.

tradallo-reputation

tradallo-reputation

MCP server for the Tradallo Verified Record Protocol — query cryptographically-verified human and AI-agent trading reputations (track records, agent version history, paginated UTRs, on-chain Solana memo notarization). Every response is JCS-canonicalized + ed25519-verified locally before returning.

md-feedback

md-feedback

An MCP server for reviewing markdown plans before AI agents implement them. Enables annotation of plans with Fix, Question, and Highlight, which AI agents can read directly through MCP.

APIbase

APIbase

Unified MCP gateway for AI agents with 56+ tools and growing. Travel (Amadeus, Sabre GDS), e-commerce, local services, financial markets (Polymarket), and marketing APIs — all through a single endpoint. Pay-per-call via x402 micropayments in USDC.

Cisco ACI Intelligent Explorer

Cisco ACI Intelligent Explorer

An MCP server that enables AI assistants to query and explore Cisco ACI fabrics using natural language by translating questions into APIC REST API calls.

HR-assist MCP server

HR-assist MCP server

Enables Claude Desktop to autonomously perform HR operations such as employee onboarding, leave management, meeting scheduling, ticket creation, and email notifications through MCP tool calling.

Granola MCP Server

Granola MCP Server

Provides tools for accessing Granola meeting transcripts and AI-generated notes, including listing recent documents, retrieving notes/transcripts, and searching by title.

knossos

knossos

MCP server for remote filesystem and CLI access over LAN, with authentication and access control via token and allowlists.

mcp-memory

mcp-memory

Enables MCP clients like Cursor and Claude to remember user preferences and information across conversations using vector search.

backtester-mcp

backtester-mcp

Local-first backtesting engine with built-in overfitting detection (PBO, deflated Sharpe, bootstrap CI, walk-forward) and a native MCP server for AI agents to validate trading strategies.

MCP Azure DevOps Server

MCP Azure DevOps Server

Enables interaction with Azure DevOps for managing work items, pull requests, and pipelines through natural language.

pagerduty-mcp

pagerduty-mcp

Enables management of PagerDuty incidents, services, schedules, and more directly from MCP-enabled clients, with embedded interactive UIs for incident command center, on-call management, and other features.

mcp-server-demo

mcp-server-demo

A demo MCP server with tools for getting weather via wttr.in and executing read-only SQLite queries.

hexforge-mcp

hexforge-mcp

A hex computation MCP server for binary security and reverse engineering, providing precise arithmetic, bitwise operations, and data conversion tools.

fallfuneral-mcp

fallfuneral-mcp

Sovereign, MIT-licensed MCP server for professional-service workflows with Ed25519-signed manifests and offline capability.

ELC Partnership Builder

ELC Partnership Builder

Enables users to build tailored company partnerships with the Engineering Leaders Community directly from their AI assistant, including qualifying goals, matching packages, customizing priced line items, and laying out 12-month journeys with an automatically applied 16% discount.

MarecoX MCP Servers

MarecoX MCP Servers

ContextForge MCP Server

ContextForge MCP Server

Enables AI agents to efficiently retrieve relevant code context via PageRank-optimized subgraphs and automate spec generation, implementation planning, and pre-commit validation.

MCP SSE Sample

MCP SSE Sample

好的,这是 MCP 服务器的 SSE(Server-Sent Events)实现的示例,包含代码和解释: **概念解释** * **SSE (Server-Sent Events):** 一种服务器向客户端推送数据的单向通信协议。客户端通过 HTTP 连接到服务器,服务器可以随时向客户端发送更新,而无需客户端发起新的请求。这非常适合实时更新,例如股票行情、新闻提要、聊天应用等。 * **MCP Server:** 我假设你指的是一个基于 Minecraft 协议 (MCP) 的服务器。虽然 MCP 本身不直接涉及 SSE,但你可以将 SSE 集成到你的服务器中,以向连接的客户端发送游戏状态或其他信息。 **示例代码 (Python + Flask)** 这个例子使用 Python 和 Flask 框架来创建一个简单的 SSE 服务器。 ```python from flask import Flask, Response, render_template import time import random app = Flask(__name__) # 模拟游戏状态数据 game_state = { "player_count": 0, "server_load": 0.0, "latest_news": "Server is online!" } def update_game_state(): """模拟更新游戏状态""" global game_state game_state["player_count"] = random.randint(0, 100) game_state["server_load"] = round(random.uniform(0.0, 1.0), 2) game_state["latest_news"] = f"Player joined! (Current: {game_state['player_count']})" def event_stream(): """生成 SSE 事件流""" while True: update_game_state() yield f"data: {game_state}\n\n" # 构建 SSE 格式的数据 time.sleep(1) # 每秒更新一次 @app.route('/') def index(): return render_template('index.html') # 渲染一个简单的 HTML 页面 @app.route('/stream') def stream(): return Response(event_stream(), mimetype="text/event-stream") if __name__ == '__main__': app.run(debug=True) ``` **解释:** 1. **导入必要的库:** * `flask`: 用于创建 Web 应用。 * `Response`: 用于构建 SSE 响应。 * `time`: 用于控制更新频率。 * `random`: 用于模拟游戏状态变化。 2. **`game_state` 字典:** * 存储模拟的游戏状态数据。你可以根据你的 MCP 服务器的需求修改这些数据。 3. **`update_game_state()` 函数:** * 模拟更新游戏状态。 在实际应用中,你需要从你的 MCP 服务器获取真实的数据。 4. **`event_stream()` 函数:** * **关键部分:** 这是一个生成器函数,它无限循环并产生 SSE 事件。 * `yield f"data: {game_state}\n\n"`: 这行代码构建了 SSE 格式的数据。 * `data:` 是 SSE 协议要求的字段,表示要发送的数据。 * `{game_state}`: 将 `game_state` 字典转换为字符串。 **重要:** 你可能需要使用 `json.dumps(game_state)` 将字典转换为 JSON 字符串,以便客户端更容易解析。 * `\n\n`: 两个换行符表示一个 SSE 事件的结束。 * `time.sleep(1)`: 暂停 1 秒,控制更新频率。 5. **`@app.route('/stream')` 路由:** * 当客户端访问 `/stream` 路径时,这个路由会被调用。 * `Response(event_stream(), mimetype="text/event-stream")`: 创建一个 `Response` 对象,将 `event_stream()` 生成器作为数据源,并将 `mimetype` 设置为 `text/event-stream`。 **`text/event-stream` 是 SSE 协议要求的 MIME 类型。** 6. **`index()` 函数和 `index.html` (可选):** * 提供一个简单的 HTML 页面,用于测试 SSE 连接。 **客户端代码 (JavaScript)** ```html <!DOCTYPE html> <html> <head> <title>SSE Example</title> </head> <body> <h1>Game State</h1> <div id="game-state"></div> <script> var eventSource = new EventSource('/stream'); // 连接到 SSE 端点 eventSource.onmessage = function(event) { var gameState = JSON.parse(event.data); // 解析 JSON 数据 document.getElementById('game-state').innerText = JSON.stringify(gameState, null, 2); // 显示格式化的 JSON }; eventSource.onerror = function(error) { console.error("SSE error:", error); }; </script> </body> </html> ``` **解释:** 1. **`new EventSource('/stream')`:** 创建一个 `EventSource` 对象,连接到服务器的 `/stream` 端点。 2. **`eventSource.onmessage`:** 定义一个事件处理函数,当服务器发送新数据时,这个函数会被调用。 * `event.data`: 包含服务器发送的数据。 * `JSON.parse(event.data)`: 将 JSON 字符串解析为 JavaScript 对象。 * `document.getElementById('game-state').innerText = JSON.stringify(gameState, null, 2)`: 将游戏状态数据显示在页面上。 `JSON.stringify(gameState, null, 2)` 用于格式化 JSON 输出,使其更易于阅读。 3. **`eventSource.onerror`:** 定义一个错误处理函数,当发生错误时,这个函数会被调用。 **如何运行:** 1. **安装 Flask:** `pip install flask` 2. **保存代码:** 将 Python 代码保存为 `app.py`,将 HTML 代码保存为 `templates/index.html` (需要在 `app.py` 所在的目录下创建一个名为 `templates` 的文件夹)。 3. **运行服务器:** `python app.py` 4. **在浏览器中打开:** `http://127.0.0.1:5000/` **重要注意事项:** * **JSON 序列化:** 在实际应用中,强烈建议使用 `json.dumps()` 将 Python 字典转换为 JSON 字符串,以便客户端更容易解析。 修改 `event_stream()` 函数如下: ```python import json def event_stream(): while True: update_game_state() yield f"data: {json.dumps(game_state)}\n\n" time.sleep(1) ``` * **错误处理:** 在客户端和服务器端添加适当的错误处理代码。 * **数据格式:** 根据你的 MCP 服务器的需求,调整 `game_state` 字典中的数据。 * **身份验证:** 如果需要,可以添加身份验证机制来保护 SSE 端点。 * **性能:** 对于高并发的场景,可能需要考虑使用更高效的异步框架,例如 `asyncio` 和 `aiohttp`。 * **MCP 集成:** 将 `update_game_state()` 函数修改为从你的 MCP 服务器获取真实的游戏状态数据。 这可能需要你使用 MCP 协议库来与服务器通信。 **总结:** 这个例子提供了一个基本的 SSE 实现,你可以根据你的 MCP 服务器的需求进行修改和扩展。 记住要使用 JSON 序列化,添加错误处理,并根据你的数据格式进行调整。 希望这个例子能帮助你理解如何在 MCP 服务器中实现 SSE。 **中文总结:** 这个例子展示了如何使用 Python 和 Flask 创建一个简单的服务器推送事件 (SSE) 服务器。 服务器模拟游戏状态数据,并通过 `/stream` 端点以 SSE 格式发送给客户端。 客户端使用 JavaScript 的 `EventSource` API 连接到服务器,接收并显示游戏状态数据。 关键点包括:使用 `text/event-stream` 作为 MIME 类型,使用 `data:` 字段格式化 SSE 数据,以及使用 `json.dumps()` 将 Python 字典转换为 JSON 字符串。 你需要根据你的 MCP 服务器的实际情况修改代码,例如从 MCP 服务器获取真实数据,并添加错误处理和身份验证。

FieldCure PublicData.Kr

FieldCure PublicData.Kr

Korean public data API gateway that enables searching, inspecting, and calling 80,000+ data.go.kr APIs (weather, real estate, air quality, etc.) via natural language.

koncepto

koncepto

Semantic concept graph MCP server for codebases — what your code means, not just what it does.

WageAPI

WageAPI

US + EU salary benchmarking, pay transparency compliance, and semantic endpoints. 1,400+ US occupations, 28 EU countries. MCP server for AI agents.

Clueso MCP

Clueso MCP

Clueso's MCP connects your favorite AI agents to a video creation engine. Just describe what you need — and every output stays fully editable, by you or AI.

LedgerLink MCP

LedgerLink MCP

Connects AI assistants to QuickBooks Online, enabling management of invoices, customers, expenses, and reports through natural language.

Playwright MCP Server

Playwright MCP Server

A Model Context Protocol server that enables AI assistants to interact with web pages through browser automation, supporting web scraping, form filling, navigation, and other browser-based tasks using Playwright.

Bun Database MCP Server

Bun Database MCP Server

A high-performance MCP server that enables AI assistants to safely interact with MySQL databases through secure CRUD operations, schema inspection, and parameterized queries with built-in SQL injection prevention.

CEDAR MCP Server

CEDAR MCP Server

Enables interaction with the CEDAR (Center for Expanded Data Annotation and Retrieval) metadata repository to fetch templates and retrieve template instances. Supports querying structured metadata and biomedical data annotations through the CEDAR platform.

superset-mcp

superset-mcp

A Model Context Protocol (MCP) server for managing Apache Superset datasets, metrics, and SQL queries.