发现优秀的 MCP 服务器

visitproject

Converts legacy supply chain databases and file drops into MCP servers for AI agents, enabling direct interaction with closed systems via CLI.

Apollo.io MCP Server

将 Apollo.io API 功能作为工具公开的 MCP 服务器

meok-mcp-injection-scan-mcp

Scans MCP servers for prompt-injection, tool-poisoning, and SSRF vulnerabilities using 30+ canonical rules across 5 severity tiers, with optional signed safety reports for procurement.

Bitrix24 MCP Server

Enables AI agents to interact with Bitrix24 CRM through comprehensive tools for managing contacts, deals, leads, companies, tasks, and users, with advanced filtering, search capabilities, and sales team performance monitoring.

MCP Server Deployment Demo

A demonstration MCP server that provides a simple addition tool for learning how to create and deploy servers following the Model-Context-Protocol specification. Serves as a basic example for developers getting started with MCP server development.

mol-open-data-mcp

Enables querying Taiwan's Ministry of Labor open data, including lists of companies violating labor laws with fuzzy search, and searching/querying ministry datasets.

qodercli-mcp

Wraps the local Qoder CLI into an MCP server, enabling any MCP client to delegate coding tasks to a local Qoder agent with session management, sandboxed permissions, and model selection.

Bitbucket MCP

A Model Context Protocol server that enables AI assistants to interact with Bitbucket repositories, pull requests, and other resources through Bitbucket Cloud and Server APIs.

SchemaCrawler AI MCP Server

Enables AI agents to interactively interrogate database schema metadata using SchemaCrawler.

STRING MCP Server

Provides access to the STRING protein-protein interaction database for mapping identifiers, retrieving interaction networks, and performing functional enrichment analysis. It enables users to explore protein partners, pathways, and cross-species homology through natural language interactions.

Deep Research MCP Server

通过整合人工智能代理、搜索引擎、网络爬虫和大型语言模型，实现迭代式的深度研究，从而高效地收集数据并生成全面的报告。

Teleprompter

Enables storage and reuse of prompt templates with variable substitution for LLMs. Supports creating, searching, and retrieving prompt templates to avoid repeating complex instructions across conversations.

scandomain-mcp

MCP server that enables AI agents to search domain, IP, port, service, title, ICP, server header, and component assets via the SCANDOMAIN API using Quake-style DSL queries.

Anaplan MCP

An MCP server that connects AI assistants to Anaplan's Integration API v2, enabling users to browse workspaces, manage model data, and execute bulk operations like imports and exports. It provides 25 structured tools to navigate model hierarchies and perform transactional tasks through natural language.

utol-mcp

Enables users to interact with the University of Tokyo's UTOL learning management system through natural language, providing access to courses, assignments, deadlines, notifications, and materials via MCP clients like Claude Desktop.

mundane-mcp

Exposes the Mundane agent-to-human marketplace as MCP tools, enabling agents to post tasks, search workers, and make offers.

Junction41 MCP Server

MCP server for the Junction41 platform, providing 125 tools for agent lifecycle, jobs, workspace, payments, bounties, and more, enabling LLMs to interact with Junction41.

cwtwb

ai generate tableau file

printify-mcp

Enables users to connect Printify to Claude and manage print-on-demand products directly from chat, including creating, updating, and publishing products, listing shops, uploading designs, and checking orders.

MCP Voice Notification

Provides voice notifications using Grok's text-to-speech API to alert users when Claude Code completes tasks, with support for both local and remote server configurations.

Scanpy-MCP

Provides a natural language interface for scRNA-Seq analysis using the Scanpy library, supporting operations such as data preprocessing, clustering, and visualization. It enables AI agents and clients to perform complex single-cell transcriptomics workflows through the Model Context Protocol.

ModelRouter MCP

Advisory MCP server for intelligent LLM model recommendations in Cursor. Provides tools for task analysis, repository analysis, and model selection guidance.

MCP-BOS

一个模块化、可扩展的模型上下文协议服务器框架，专为 Claude Desktop 设计，它采用基于约定的自动模块发现机制，以便在不修改核心代码的情况下轻松扩展 AI 应用程序的功能。

zuul-mcp

MCP server for Zuul CI/CD with 25 tools for builds, pipelines, queue management (enqueue/dequeue/promote), infrastructure visibility, and autohold management. Supports stdio, HTTP, and SSE transports.

Marsa Atlas

MCP server providing source-grounded migration services and evidence for Egypt, with tools for search, service discovery, eligibility checks, and research briefs.

sesha-skills

A stdio MCP server providing everyday tools and Markdown skills for the Sesha agent, enabling notes, web research, coding, docs, scheduling, and safety/governance.

mdapi-mcp

MCP server for converting HTML to Markdown and extracting page data, with pay-per-call using USDC on Base.

fleet-mcp

A unified MCP server for managing hosting fleets, enabling natural language control over SSH, WordPress, Cloudflare, MySQL, GitHub, Docker, Coolify, and more.

mcp-talib

一个提供 ta-lib-python 功能的 Model Context Protocol (MCP) 服务器。 Or, more literally: 提供 ta-lib-python 功能的 Model Context Protocol (MCP) 服务器。 (Tígōng ta-lib-python gōngnéng de Model Context Protocol (MCP) fúwùqì.)

MCP-llms-txt

Okay, I understand. You want me to: 1. Create an MCP (presumably referring to a Minecraft Protocol) server. 2. This server should be related to the project "SecretiveShell/Awesome-llms-txt". 3. I should add documentation directly into our conversation, using MCP resources (presumably meaning Minecraft Protocol resources, like packets and data structures). This is a complex request that requires significant coding and understanding of Minecraft's internal workings. I can't *actually* create and host a server for you. That requires a development environment, a Minecraft server instance, and the ability to write and execute code. However, I *can* provide you with a conceptual outline and code snippets to get you started, along with documentation integrated into our conversation. I'll focus on the core aspects of handling a connection and sending/receiving basic data. **Conceptual Outline** 1. **Server Setup:** Use a programming language like Java (the language Minecraft is written in) or Python (with a library like `mcstatus` or `nbt`) to create a server socket that listens for incoming connections on a specific port (e.g., 25565, the default Minecraft port). 2. **Handshake:** The Minecraft client initiates a handshake. You need to parse this handshake packet to determine the protocol version and the intended server state (status or login). 3. **Status/Login:** * **Status:** If the client requests status, you send back a JSON response containing server information (MOTD, player count, etc.). * **Login:** If the client requests login, you handle authentication (if required) and then transition the client to the play state. 4. **Play State:** This is where the core game logic happens. You receive packets from the client (e.g., movement, chat messages) and send packets back to the client (e.g., world updates, entity positions). **Simplified Code Snippet (Python using `socket` - for demonstration only, not a full MCP implementation):** ```python import socket import json HOST = '127.0.0.1' # Standard loopback interface address (localhost) PORT = 25565 # Minecraft default port with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s: s.bind((HOST, PORT)) s.listen() print(f"Server listening on {HOST}:{PORT}") conn, addr = s.accept() with conn: print(f"Connected by {addr}") while True: data = conn.recv(1024) # Receive up to 1024 bytes if not data: break print(f"Received: {data}") # **VERY SIMPLIFIED HANDSHAKE EXAMPLE (DOES NOT PARSE PROPERLY)** if b'\x00\x04' in data: # Crude check for a handshake-like packet print("Possible Handshake detected") # **IN REALITY, YOU NEED TO PARSE THE VARINTS AND DATA PROPERLY** # Example Status Response (Simplified) status_response = { "version": {"name": "My Awesome Server", "protocol": 757}, "players": {"max": 100, "online": 10, "sample": []}, "description": {"text": "A server for Awesome-llms-txt!"} } json_response = json.dumps(status_response) # **IMPORTANT: Minecraft requires a VarInt length prefix before the JSON** # **This is a placeholder - you need to implement VarInt encoding** length_prefix = len(json_response).to_bytes(1, 'big') # Incorrect VarInt encoding conn.sendall(length_prefix + json_response.encode('utf-8')) else: conn.sendall(b"Received your data!") # Echo back (for testing) ``` **Explanation and MCP Documentation Integration** * **`socket.socket(socket.AF_INET, socket.SOCK_STREAM)`:** Creates a TCP socket. TCP is the protocol Minecraft uses. This corresponds to the underlying network layer. * **`s.bind((HOST, PORT))`:** Binds the socket to a specific IP address and port. * **`s.listen()`:** Starts listening for incoming connections. * **`conn, addr = s.accept()`:** Accepts a connection. `conn` is a new socket object for communicating with the client, and `addr` is the client's address. * **`conn.recv(1024)`:** Receives data from the client. The `1024` is the maximum number of bytes to receive at once. * **Handshake (MCP Relevant):** The handshake is the first packet sent by the client. It contains: * **Packet ID (VarInt):** `0x00` for handshake. * **Protocol Version (VarInt):** The Minecraft protocol version the client is using. This is *crucial* for compatibility. 757 is for 1.17.1. You *must* handle different protocol versions. * **Server Address (String):** The address the client connected to. * **Server Port (Unsigned Short):** The port the client connected to. * **Next State (VarInt):** `1` for status, `2` for login. **VarInt Encoding:** A VarInt is a variable-length integer. It uses one or more bytes to represent an integer. Each byte (except the last) has its most significant bit set to 1. The lower 7 bits of each byte are used to store the integer's value. This is a *critical* part of the Minecraft protocol. The example code *incorrectly* uses `len(json_response).to_bytes(1, 'big')` which is *not* a VarInt. You need a proper VarInt encoding function. ```python def encode_varint(number): buf = [] while True: byte = number & 0x7F # Get the lowest 7 bits number >>= 7 if number: byte |= 0x80 # Set the MSB to indicate more bytes buf.append(byte) if not number: break return bytes(buf) ``` * **Status Response (MCP Relevant):** The status response is a JSON string that contains server information. The JSON is *prefixed* with a VarInt indicating the length of the JSON string. The structure of the JSON is defined by the Minecraft protocol. The example code provides a simplified version. **Next Steps and Considerations** 1. **VarInt Implementation:** Implement proper VarInt encoding and decoding. This is essential for handling all packets. 2. **Packet Parsing:** Implement proper packet parsing based on the protocol version. Use a library or write your own code to read VarInts, strings, and other data types from the byte stream. 3. **Protocol Version Handling:** Support multiple Minecraft protocol versions. This is a *major* undertaking, as the protocol changes frequently. You'll need to maintain a mapping of protocol versions to packet structures. 4. **Authentication:** Implement authentication if you want to require players to log in with a Minecraft account. This involves interacting with Mojang's authentication servers. 5. **Game Logic:** Implement the core game logic for your server. This will involve handling player movement, world updates, and other game events. 6. **NBT Data:** Minecraft uses Named Binary Tag (NBT) format for storing world data, player data, and other complex data structures. You'll need a library to read and write NBT data. 7. **Asynchronous Handling:** Use asynchronous programming (e.g., `asyncio` in Python) to handle multiple clients concurrently. This is a very high-level overview. Building a Minecraft server from scratch is a significant project. Start with the basics (handshake and status) and gradually add more features. Good luck! Let me know if you have more specific questions. I can provide more detailed code snippets and explanations for specific parts of the protocol.