DEV Community

Wanda
Wanda

Posted on • Originally published at roobia.Medium on

The 10 Must-Have MCP Servers for Claude Code (2025 Developer Edition)

Modern developers crave tools that automate, integrate, and accelerate their coding experience. Claude Code, Anthropic’s command-line AI assistant, is powerful on its own — but when you connect it to Model Context Protocol (MCP) servers, it becomes a true coding powerhouse. MCP servers act as bridges, letting Claude Code interact with external APIs, databases, file systems, and more, all in real time. The result? You can automate tasks, access live data, and streamline your workflow — without ever leaving your terminal.

What is MCP and Why Does It Matter?

Model Context Protocol (MCP) is Anthropic’s open standard for connecting AI models like Claude to external tools. Think of MCP as a universal adapter: it lets Claude Code “talk” to servers that expose specific functions, from file operations to API calls. This modular approach means you can customize your AI assistant’s capabilities with plug-and-play ease.

MCP has three main parts:

  • Host: The app (Claude Code, Cursor, etc.) that makes requests
  • Client: The go-between for host and servers
  • Server: The tool or service (like GitHub, Apidog, or your local file system)

With MCP, Claude Code isn’t just a chatbot — it’s a full-featured automation and productivity engine. Here are the top 10 MCP servers every developer should know in 2025.

1. GitHub MCP Server: Version Control, Automated

GitHub MCP server connects Claude Code to GitHub’s API. Read issues, manage PRs, trigger CI/CD, and analyze commits — all from your terminal.

Why use it?

  • Automate GitHub tasks (comment, merge, etc.)
  • Pull commit histories for debugging
  • Stay focused — no more context switching

Setup:

  • Install Node.js, run npm install @composio/mcp@latest
  • npx @composio/mcp@latest setup github --client claude
  • Authenticate via OAuth in Claude Code’s settings
  • Restart Claude Code

Example:

Ask Claude Code to “find all issues about authentication” — get instant results.

2. Apidog MCP Server: Connect AI to API Specifications

Apidog MCP server bridges the gap between your API specifications and AI-powered development workflows. This essential MCP server allows Claude and other AI assistants to directly access and work with your API documentation, making code generation more accurate and contextually aware.

Apidog MCP Server: The Bridge Between Your API Specifications and AI Coding Assistants

Key Capabilities

  • API Specification Integration: Connects AI to your Apidog project, online API documentation published by Apidog, or local OpenAPI/Swagger files
  • Intelligent Code Generation: Generate DTOs, controllers, and client code based on your actual API contracts
  • Real-time Updates: AI can refresh cached API data to work with the latest specification changes

Why Developers Love It

The Apidog MCP Server eliminates the guesswork from API-driven development. Instead of manually translating API specs into code, developers can simply ask Claude to “generate Java records for the Product schema” or “add the new fields to the User DTO based on the API specification.” The AI pulls directly from your authoritative API documentation, ensuring generated code matches your exact contracts.

Setup Process

Let’s say, now we want to use local OpenAPI/Swagger files as the data source for Claude AI.

  1. Open Claude Code’s settings and navigate to the MCP tab.
  2. Add the Apidog MCP server configuration to mcp.json:
//for macOS/Linux: { "mcpServers": { "API specification": { "command": "npx", "args": [ "-y", "apidog-mcp-server@latest", "--oas=<oas-url-or-path>" ] } } } //for Windows: { "mcpServers": { "API specification": { "command": "cmd", "args": [ "/c", "npx", "-y", "apidog-mcp-server@latest", "--oas=<oas-url-or-path>" ] } } } 
Enter fullscreen mode Exit fullscreen mode
  1. Test the connection by asking Claude Code to fetch one of the endpoint specifications.

  2. If the AI successfully returns information that meets the expectations, the connection is established!

3. File System MCP Server: Local File Power

This server lets Claude Code read, write, and edit files on your machine. Perfect for project management, log analysis, or quick edits.

Why use it?

  • Automate file operations (CRUD)
  • Keep project context at your fingertips
  • Refactor or clean up files with a prompt

Setup:

Example:

“Update the README.md with a new section” — done in seconds.

4. Sequential Thinking MCP Server: Smarter Problem Solving

This server helps Claude Code break down complex tasks into logical steps — great for architecture, refactoring, or planning.

Why use it?

  • Guides Claude Code to think step-by-step
  • Handles big projects with clear logic
  • Makes debugging and design easier

Setup:

  • npm install -g @modelcontextprotocol/server-sequential-thinking
  • Add to claude_desktop_config.json:
{ "mcpServers": { "sequential-thinking": { "command": "node", "args": ["sequential-thinking.js"] } }} 
Enter fullscreen mode Exit fullscreen mode
  • Restart and test with “Break down the steps to refactor this module.”

Example:

“Outline the steps to decouple this service” — get a detailed plan.

5. Puppeteer MCP Server: Web Automation

Control browsers for scraping, testing, or automating workflows. Navigate pages, take screenshots, and interact with web elements.

Why use it?

  • Automate browser tasks
  • Run UI tests
  • Scrape web data for analysis

Setup:

  • npm install puppeteer
  • Clone the repo, install dependencies
  • Configure Claude Code and restart

Example:

“Take a screenshot of this webpage” — get the image instantly.

6. PostgreSQL MCP Server: Natural Language Database Queries

Query your database using plain English. No more manual SQL — just ask and get results.

Why use it?

  • Translate natural language to SQL
  • Retrieve and manipulate data easily
  • Boost productivity for non-SQL experts

Setup:

Example:

“Summarize sales data from the past month” — get a formatted report.

7. Notion MCP Server: Docs and Tasks, Synced

Connect Claude Code to Notion for instant access to docs, tasks, and project specs.

Why use it?

  • Update Notion tasks from the terminal
  • Fetch project specs for context
  • Sync team workflows with AI

Setup:

  • npx @composio/mcp@latest setup notion --client claude
  • Authenticate via OAuth

Example:

“Add a new task to Notion for code review” — no browser needed.

8. Memory Bank MCP Server: Persistent Context

Give Claude Code a memory! Retain context across sessions, perfect for big projects or long-term tracking.

Why use it?

  • Recall prior interactions
  • Maintain coherence in large codebases
  • Reduce repetitive explanations

Setup:

Example:

“Resume work on the last module I edited” — Claude Code remembers.

9. Figma MCP Server: Design-to-Code Magic

Bridge the gap between design and code. Convert Figma layouts into code snippets or UI components.

Why use it?

  • Turn Figma designs into code
  • Rapid prototyping
  • Align devs and designers

Setup:

  • npx @composio/mcp@latest setup figma --client claude
  • Authenticate via OAuth

Example:

“Convert this Figma layout to React components” — get production-ready code.

10. Zapier MCP Server: Automate Everything

Connect Claude Code to Zapier for cross-app automation — trigger actions in Slack, Gmail, Trello, and more.

Why use it?

  • Automate multi-app workflows
  • Streamline notifications and updates
  • Integrate with hundreds of services

Setup:

  • npx @composio/mcp@latest setup zapier --client claude
  • Authenticate with Zapier

Example:

“Send a Slack message when a new PR is opened” — team communication, automated.

How to Choose the Right MCP Server

  • Task type: Web automation? Use Puppeteer. Database work? PostgreSQL.
  • Setup: Look for clear docs and OAuth support (Apidog, Notion).
  • Scale: Use Memory Bank for big projects, Sequential Thinking for complex logic.
  • Integration: Zapier for cross-app, GitHub for version control.

Test servers with simple prompts, and always secure sensitive data by restricting access.

Tips for Getting the Most from MCP Servers

  • Keep configs clean — avoid typos in claude_desktop_config.json
  • Use debug flags (--mcp-debug) to troubleshoot
  • Store prompt templates in .claude/commands
  • Combine servers for powerful workflows
  • Limit active servers to avoid slowdowns

Why MCP Servers Are a Game-Changer for Claude Code

MCP servers turn Claude Code into a true AI-powered development assistant. By connecting to tools like GitHub, Apidog, and PostgreSQL, you can automate, integrate, and accelerate your workflow. The modular design means new servers are always being developed, so your setup stays future-proof.

Ready to level up your coding? Start experimenting with these top MCP servers — and don’t forget to download Apidog for free to make your API-driven projects even smoother.

Top comments (0)