Install commands are being verified. Only servers with the Verified badge have been checked against the real package registry. Unverified commands may 404 -- always inspect the source repo before running. Submit corrections on GitHub.
10 curated servers
AI engineering requires orchestrating multiple models, retrieval systems, and external data sources. This starter kit collects the MCP servers that make building AI applications faster. Use memory servers for persistent context across conversations, RAG servers for retrieval-augmented generation, web search via Tavily and Perplexity for grounding answers in current data, code execution in sandboxed E2B environments, and multi-model routing through OpenAI and Hugging Face. Whether you are building agents, chatbots, or AI-augmented workflows, these servers provide the building blocks.
by Anthropic
Knowledge graph-based persistent memory system. Enables Claude to remember information across conversations using entities and relations.
npx -y @modelcontextprotocol/server-memoryby Anthropic
Dynamic and reflective problem-solving through thought sequences. Helps AI models break down complex problems step by step.
npx -y @modelcontextprotocol/server-sequential-thinkingby nicepkg
Retrieval-Augmented Generation server. Index documents, perform semantic search, and build RAG pipelines.
npx -y rag-mcp-serverby Tavily
AI-powered web search with Tavily. Get search results optimized for LLMs with clean, relevant content extraction.
npx -y tavily-mcpby nicepkg
Access Perplexity AI for research queries. Get cited answers with source references for any topic.
npx -y perplexity-mcpby E2B
Run code in secure E2B cloud sandboxes. Execute Python, JavaScript, and other languages in isolated environments.
npx -y e2b-mcp-serverby nicepkg
Access OpenAI APIs through MCP. Use GPT models, DALL-E, Whisper, and embeddings in your AI workflows.
npx -y mcp-openaiby nicepkg
Search models, datasets, and spaces on Hugging Face. Download models, run inference, and explore the ML ecosystem.
npx -y huggingface-mcpby LangChain
Bridge between LangChain and MCP. Use MCP tools as LangChain tools and integrate with LangGraph agents.
pip install langchain-mcp-adaptersby Upstash
Fetch up-to-date documentation for any library or framework. Get accurate API references and code examples on demand.
npx -y @upstash/context7-mcpQuery databases, analyze datasets, search papers, and run experiments from your AI assistant.
10 servers
Ship faster with servers for GitHub, Vercel, databases, browser testing, and error tracking.
10 servers
Manage infrastructure, containers, CI/CD pipelines, and monitoring from your terminal.
10 servers