MCPALL
MCPALL es un directorio de servidores MCP (categoría: Directorio/Recurso) con guías originales, comandos de instalación y configuración de clientes para Claude, Cursor, VS Code y Windsurf.
Documentación
Popular MCP Servers
BlenderMCP
Games & 3D
BlenderMCP links AI assistants to Blender for scene inspection, object and material changes, and Python execution. Optional asset integrations add Poly Haven downloads, Sketchfab models, and generated 3D assets.
Notion MCP Server
Memory & Knowledge
Connect an MCP client to the Notion API to work with pages and data sources. This local server supports Markdown content operations, but Notion now prioritizes its separate remote MCP service.
Slack MCP Server
Communication
Connect an assistant to Slack channels, threads, DMs, and user information. The server supports filtered history and search, multiple transports, and optional posting or reactions disabled by default.
Playwright MCP
Browser Automation
Playwright MCP lets AI assistants operate browsers using structured accessibility data. It suits agents that need persistent page context, interactive exploration, and browser inspection without a vision model.
Firecrawl MCP Server
Search & Web Data
Firecrawl MCP connects AI agents to web search, scraping, crawling, and live browser interaction. It offers hosted access or a local server for collecting structured web context and running research tasks.
Framelink MCP for Figma
Design & Creative
Framelink supplies coding agents with selected Figma metadata and image assets. Developers can bring a design link into an editor conversation and use its underlying layout details while implementing an interface.
21st.dev Magic AI Agent
Design & Creative
21st.dev Magic AI Agent turns natural-language requests into customizable UI components inside supported coding tools. It combines component-library inspiration, previews, TypeScript support, and brand assets for interface development.
MiniMax MCP
AI Models & Agents
MiniMax MCP exposes MiniMax media APIs to compatible AI clients for speech, voice work, images, video, and music. It supports local stdio use and SSE deployment, with region-specific API credentials and configurable resource output.
MCP Servers by Category
[
Browser Automation
11
Browser automation gives an assistant a way to interact with websites rather than merely describe them. Servers in this category can expose browser sessions, page inspection, navigation, or scripted interactions through MCP.
](https://mcpall.com/category/browser-automation)[
Search and web data servers connect an assistant to information beyond its conversation history. They may provide search results, retrieve a selected webpage, or extract structured material from an online source.
](https://mcpall.com/category/search-web-data)[
Developer Tools
18
Developer tool servers bring software engineering context and actions into an MCP client. Depending on the project, that might mean inspecting code, consulting framework documentation, running a development utility, or examining a debugging session.
](https://mcpall.com/category/developer-tools)[
Databases
18
Database integrations let an assistant work with structured records through a defined tool interface. Some servers specialize in a particular database engine, while others sit in front of a hosted data service.
](https://mcpall.com/category/databases)[
Cloud and DevOps servers connect assistants to the systems that run applications. Their scope can range from reading deployment status to interacting with infrastructure APIs or inspecting operational information.
](https://mcpall.com/category/cloud-devops)[
Memory and knowledge servers help assistants revisit information across tasks. A project may store notes, expose a knowledge graph, search a document collection, or retrieve context from an existing knowledge service.
](https://mcpall.com/category/memory-knowledge)[
AI model and agent servers make another layer of intelligence available through MCP. They may connect a client to a model provider, a specialized generation service, or a workflow that coordinates multiple reasoning steps.
](https://mcpall.com/category/ai-models-agents)[
Data and analytics servers help an assistant move from a business question to supporting measurements. An integration might expose a reporting platform, a metric service, or tools for examining datasets.
](https://mcpall.com/category/data-analytics)[
Design and creative integrations connect assistants to tools used for visual communication and content production. A server might inspect a design document, access an asset library, or support a task inside a creative application.
](https://mcpall.com/category/design-creative)[
Communication
10
Communication servers connect assistants to the places where people exchange information. Depending on the integration, a client may be able to find messages, read conversation history, prepare a response, or send a notification.
](https://mcpall.com/category/communication)[
Productivity
15
Productivity integrations bring an assistant closer to the systems used to organize everyday work. They can connect to planning tools, task trackers, calendars, or other workspace services, with the available actions determined by each project.
](https://mcpall.com/category/productivity)[
Files and document servers give assistants access to material stored outside the conversation. A project may expose a local directory, connect to a document platform, or read a particular file format.
](https://mcpall.com/category/files-documents)[
Games and 3D integrations connect assistants to environments where scenes, objects, and interactive behavior are built. A server might communicate with an editor, inspect a project, or expose operations in a modeling application.
](https://mcpall.com/category/game-3d)[
Security and reverse engineering servers bring specialized inspection tools into an assistant's workflow. They may help examine software artifacts, navigate analysis results, or interact with a security application.
](https://mcpall.com/category/security-reverse-engineering)[
Gateways and proxies sit between an MCP client and other tool services. They can address connection management, transport adaptation, or access to multiple endpoints, depending on the project.
](https://mcpall.com/category/mcp-gateways-proxies)[
Mobile and desktop automation servers connect assistants to interfaces and applications beyond the browser. A project may target a device, a simulator, a desktop environment, or a particular operating system.
](https://mcpall.com/category/mobile-desktop-automation)[
Finance and commerce servers connect assistants to financial information or the systems behind buying and selling. Some integrations retrieve public market data, while others expose account records or commerce platform operations.
](https://mcpall.com/category/finance-commerce)[
Maps and travel integrations connect assistants to information about places and journeys. A server may search for locations, retrieve geographic details, or support a planning workflow through a travel service.
](https://mcpall.com/category/maps-travel)[
Blockchain and Web3 integrations give an assistant access to networks and services whose records are organized around addresses, transactions, and smart contracts. Some servers retrieve public chain activity or explain a contract interface; others connect to wallets, token markets, or…
](https://mcpall.com/category/blockchain-web3)
MCP Servers by AI Client
The same server needs different configuration in different clients. Start with a guide for your editor or assistant. MCP is also supported by ChatGPT, Codex, Gemini CLI and Claude custom connectors.
Claude Desktop
Configuration, setup, and a careful first tool call.
Cursor
Configuration, setup, and a careful first tool call.
VS Code
Configuration, setup, and a careful first tool call.
Windsurf
Configuration, setup, and a careful first tool call.
Claude Code
Configuration, setup, and a careful first tool call.
Remote MCP Servers
Remote servers use Streamable HTTP and usually OAuth for authorization. The official MCP Registry (preview) lists public servers for marketplaces to build on. Browse the official MCP Registry (preview)
Full MCP Server List
Up to 20 published servers with recorded GitHub stars. Shared monorepo stars, archived repositories and non-server projects are excluded. Stars measure popularity, not security or compatibility.
- 1.Context7 63,226 stars
- 2.Playwright MCP 38,311 stars
- 3.GitHub MCP Server 33,713 stars
- 4.BlenderMCP 31,468 stars
- 5.Serena 30,351 stars
- 6.MCP Toolbox for Databases 16,678 stars
- 7.Framelink MCP for Figma 16,002 stars
- 8.MCP for Unity 15,074 stars
- 9.IDA Pro MCP 12,992 stars
- 10.claude-context 12,622 stars
- 11.FastAPI-MCP 12,023 stars
- 12.Apify MCP Server 11,782 stars
- 13.ghidraMCP 11,367 stars
- 14.Desktop Commander MCP 10,092 stars
- 15.Mobile Next Mobile MCP 9,719 stars
- 16.Windows-MCP 9,410 stars
- 17.GitMCP 8,471 stars
- 18.Firecrawl MCP Server 7,629 stars
- 19.BrowserTools MCP 7,327 stars
- 20.Browser MCP 7,193 stars
Browse all 237 MCP servers · Browse the official MCP Registry (preview)
Official MCP Servers
Reference servers with repository URLs under the modelcontextprotocol GitHub organization. Repository ownership is not a security review or a guarantee of ongoing maintenance.
What Is an MCP Server? (MCP vs API)
MCP (Model Context Protocol) is an open standard for connecting AI applications to external systems. Anthropic open-sourced it in November 2024 and donated it on December 9, 2025 to the Agentic AI Foundation (AAIF), a directed fund under the Linux Foundation co-founded by Anthropic, Block and OpenAI. MCP's maintainer-led governance continues. Servers expose tools, resources and prompts, and can ask users for input through elicitation.
The current spec revision, 2026-07-28, makes MCP stateless: there is no initialize handshake or session; each request carries version and capabilities. It adds server/discover and Multi Round-Trip Requests, allowing a server to ask for more input without a long-lived session. Roots, Sampling and Logging are deprecated.
Local servers use stdio; remote servers use Streamable HTTP. The legacy HTTP+SSE transport is deprecated; migrate to Streamable HTTP. OAuth is recommended for authorizing remote servers. Dynamic Client Registration is deprecated in favor of Client ID Metadata Documents.
The official MCP Registry is a metadata registry for publicly accessible servers with namespace verification, mainly consumed by downstream registries and marketplaces. It is in preview and may have breaking changes before general availability.
An API is a service-specific interface that each integration must code against. MCP is a standard protocol so any compatible AI client can discover and call a server's tools. Many MCP servers wrap existing APIs: MCP complements APIs rather than replacing them.
Read the MCP specification and documentation
Frequently asked questions
What is an MCP server?
An MCP server connects AI applications to external systems through the Model Context Protocol (MCP), an open standard. Servers expose tools, resources and prompts, and can ask users for input through elicitation.
How do I add an MCP server to ChatGPT, Claude, Cursor or Codex?
For clients that support local servers, add the server's command and arguments to the client's MCP configuration. For remote servers, add the server URL through the client's supported connection flow; OAuth is the recommended authorization method. ChatGPT connects to remote MCP servers via apps/plugins; full write actions are in beta for Business, Enterprise and Edu. Claude custom connectors use remote MCP. Codex's CLI, IDE extension and ChatGPT desktop app share ~/.codex/config.toml.
Are MCP servers safe?
MCP support is not a safety guarantee. Local servers run code with your privileges: review the exact command before running it and prefer sandboxing. Grant least-privilege scopes. Servers must not accept tokens that were not issued for them—no token passthrough. GitHub stars are not a security audit.
Local vs remote MCP servers?
Local MCP servers run on your machine and use stdio. Remote MCP servers are reached over the network using Streamable HTTP, with OAuth recommended for authorization. The legacy HTTP+SSE transport is deprecated; migrate to Streamable HTTP.
What is the difference between MCP and an API?
An API is a service-specific interface that each integration must code against. MCP is a standard protocol so any compatible AI client can discover and call a server's tools. Many MCP servers wrap existing APIs: MCP complements APIs rather than replacing them.