发现优秀的 MCP 服务器

Azure Pricing MCP Server

This MCP server enables querying Azure resource pricing programmatically through a structured workflow, including listing service families, service names, products, and calculating monthly costs.

swarm-suite

An AI engineering team for embedded firmware, lab automation, and hardware drivers. It uses 53 specialized experts and a datasheet-to-release pipeline to produce auditable, hardware-aware code.

HFSS MCP Server

An MCP server for automating ANSYS HFSS simulations, providing persistent connections and a suite of tools for modeling, simulation, and post-processing.

grantguard-mcp

Enables auditing PostgreSQL role write privileges against a policy file, using read-only tools to explain, describe, and check permissions.

personal-notes-assistant

Enables querying, listing, and summarizing personal knowledge base documents using RAG with hybrid search and LLM.

ai-assimilation-mcp

Enables AI models to securely and structurally assimilate experiences, thoughts, and reasoning processes across different AI systems, fostering intellectual collaboration and evolutionary dialogue.

mcp-server-wechat

好的，我理解您想了解如何在PC端微信上实现MCP（Message Communication Protocol，消息通信协议）服务功能。 这实际上是一个比较复杂的问题，涉及到逆向工程、协议分析、以及代码实现等多个方面。 **请注意：** * **法律风险：** 逆向工程和修改微信客户端可能违反微信的使用条款，甚至触犯法律。请务必在法律允许的范围内进行研究和开发。 * **技术难度：** 微信客户端是一个复杂的软件，其协议和实现细节可能会随时更改。实现MCP服务功能需要深入的理解和持续的维护。 * **安全风险：** 修改微信客户端可能会引入安全漏洞，导致您的账号或设备受到攻击。 **以下是一些可能涉及的步骤和技术方向，仅供参考：** 1. **逆向工程和协议分析：** * **目标：** 了解PC端微信如何收发消息，以及MCP协议的具体格式和流程。 * **工具：** * **反汇编器 (Disassembler):** IDA Pro, Ghidra 等，用于分析微信客户端的可执行文件。 * **调试器 (Debugger):** x64dbg, OllyDbg 等，用于动态调试微信客户端，观察其运行时的行为。 * **抓包工具 (Packet Sniffer):** Wireshark, Fiddler 等，用于捕获微信客户端与服务器之间的网络流量，分析MCP协议。 * **步骤：** * 使用反汇编器打开微信客户端的可执行文件，找到与消息收发相关的函数和代码段。 * 使用调试器跟踪这些函数的执行过程，观察其如何构建和解析MCP消息。 * 使用抓包工具捕获微信客户端与服务器之间的网络流量，分析MCP协议的格式和字段。 * 重点关注登录、消息发送、消息接收、群组管理等关键功能的协议细节。 2. **MCP协议模拟：** * **目标：** 能够模拟微信客户端，构造和发送符合MCP协议的消息，并解析接收到的消息。 * **技术：** * **编程语言：** C++, Python, Go 等，选择您熟悉的语言。 * **网络编程：** TCP/IP 协议，Socket 编程。 * **数据序列化/反序列化：** Protocol Buffers, JSON 等，根据MCP协议的格式选择合适的工具。 * **加密算法：** 微信可能使用了加密算法来保护MCP消息，需要分析并实现相应的解密算法。 * **步骤：** * 根据协议分析的结果，编写代码来构造MCP消息。 * 使用Socket编程建立与微信服务器的连接。 * 发送构造好的MCP消息，并接收服务器的响应。 * 解析接收到的MCP消息，提取所需的信息。 * 实现登录、消息发送、消息接收等基本功能。 3. **PC端微信功能扩展 (可选):** * **目标：** 将MCP服务功能集成到PC端微信客户端中，实现自定义的功能。 * **技术：** * **DLL注入 (DLL Injection):** 将自定义的DLL注入到微信客户端的进程中。 * **Hook技术 (Hooking):** 拦截微信客户端的函数调用，修改其行为。 * **内存修改 (Memory Patching):** 直接修改微信客户端的内存数据。 * **步骤：** * 编写一个DLL，实现MCP服务功能。 * 使用DLL注入工具将DLL注入到微信客户端的进程中。 * 使用Hook技术拦截微信客户端的函数调用，例如消息发送函数、消息接收函数等。 * 在Hook函数中调用DLL中的MCP服务功能，实现自定义的功能。 * 或者，直接修改微信客户端的内存数据，例如修改消息显示内容、添加自定义按钮等。 **一些建议：** * **从小处着手：** 先从简单的功能开始，例如模拟登录、发送简单的文本消息等。 * **逐步迭代：** 不断完善和扩展功能，逐步实现更复杂的功能。 * **参考开源项目：** 在GitHub等平台上搜索相关的开源项目，可以借鉴其实现思路和代码。 * **注意安全：** 在开发过程中，要注意保护您的账号和设备安全，避免泄露个人信息。 **总结：** 实现PC端微信的MCP服务功能是一个极具挑战性的任务，需要深入的理解微信客户端的内部机制和MCP协议。 请务必在法律允许的范围内进行研究和开发，并注意安全风险。 希望这些信息对您有所帮助！ 如果您有更具体的问题，例如关于某个特定功能的实现，请随时提出。

x-wing-mcp

An MCP server for X (Twitter) that enables creating posts, threads, likes, follows, and DMs, plus reading posts, timelines, mentions, and search results via a unified OAuth 2.0 connection.

trmm-mcp

MCP server that gives AI assistants read-only access to TacticalRMM fleet data (agents, checks, alerts, services, processes, etc.) and, with explicit approval, lets them run commands on managed machines.

Remote MCP Server Authless

A Cloudflare Workers-based remote Model Context Protocol server that operates without authentication requirements, allowing users to deploy custom AI tools that can be accessed from Claude Desktop or the Cloudflare AI Playground.

cloudflare-mcp-server

Enables natural language management of Cloudflare services including Workers, KV, R2, D1, Queues, and more via the Model Context Protocol.

SnackMachine

Modular snack registry, scheduler, spool, and MCP knowledge conduit that enables AI-ranked search, multi-agent shared memory, and FTS5 indexing across vault layers.

accountinghouse-mcp

MCP server for foldkit that exposes the 7-prime spine, 7 κ-bands, and 6 fold operations as native tools and resources in any MCP client.

crux-mcp

MCP server for the Chrome UX Report, providing real-user Core Web Vitals (LCP, INP, CLS) and historical trends for any origin or URL via natural language queries.

AI-Powered Kubernetes MCP Server

Enables natural-language management of Kubernetes clusters via kubectl-ai and Gemini, providing commands like listing pods, scaling deployments, and retrieving logs through a FastAPI backend.

latvian-cybersecurity-mcp

Enables querying Latvian cybersecurity data, including regulations, guidelines, and advisories from CERT.lv, directly from any MCP-compatible client.

Firecrawl MCP Server

A Model Context Protocol server that enables web scraping, crawling, and content extraction capabilities through integration with Firecrawl.

Epstein-Files-Plugin

Enables AI agents to search and query the public DOJ Epstein Files release across 20 databases, resolve EFTA numbers to official DOJ PDFs, and cross-check viral claims against a fact-checked registry.

Ratatosk

CNCF release intelligence: typed, quoted facts (security fixes, breaking changes, deprecations) from every graduated/incubating project's release notes, updated daily. check_stack compares running versions locally — versions never leave your process.

Gmail AutoAuth MCP Server

Enables AI assistants to manage Gmail through natural language interactions, supporting email operations (send, read, search, draft), comprehensive attachment handling (send, receive, download), label management, filters, and batch operations with automatic OAuth2 authentication.

whoop-mcp

Connects WHOOP fitness tracker data to AI assistants like Claude and ChatGPT, enabling natural language queries about recovery, sleep, strain, and trends.

CST Studio Orchestrator MCP

Enables AI agents to control CST Studio Suite for 3D electromagnetic simulation, antenna design, and schematic-based field-circuit co-simulation through 177 MCP tools.

Browser MCP

Enables MCP clients to control a real local browser window for web automation tasks such as clicking, typing, scrolling, and taking screenshots.

chromium-sidecar

Enables local AI agents to operate the user's existing browser session via a Manifest V3 extension and native messaging, supporting tabs, DOM snapshots, clicks, form input, JavaScript, screenshots, cookies, network capture, and Arc Spaces.

Workshop

Enables LLM clients to query current weather and temperature for a city and simulate sending invitations with confirmation and per-item progress.

Ecosyste.ms Package Data

Enables querying package ecosystem data from ecosyste.ms, including package metadata, versions, security advisories, dependencies, and repository information across 40+ package registries with fast local SQLite lookups and API fallback.

ToolForge MCP Server

Enables AI agents to securely discover, execute, and observe tools with role-based access control and audit logging. Serves tools over MCP stdio and HTTP for integration with Claude Desktop, Cursor, and other clients.

la-caja-mcp

Enables structured debates with agent-agent-human interaction and access to persistent contextual memory (La Caja) through MCP tools.

claude-code-delegate-local

MCP server that delegates Claude Code subagents to alternative backends such as local models, DeepSeek, or AWS Bedrock, while keeping your Claude Code orchestrator session intact.

mcp-browser-use

Enables browser automation through MCP clients like Claude or Cursor, using the client's existing LLM without requiring an additional API key.