Best MCP Servers for RAG Systems

claude_desktop_config.json

{
  "mcpServers": {
    "exa": {
      "command": "npx",
      "args": ["-y", "@exa/mcp-server"],
      "env": {
        "EXA_API_KEY": "your_exa_api_key_here"
      }
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
      "env": {
        "BRAVE_API_KEY": "your_brave_api_key_here"
      }
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "@firecrawl/mcp-server"],
      "env": {
        "FIRECRAWL_API_KEY": "your_firecrawl_api_key_here"
      }
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/Documents"],
      "env": {}
    }
  }
}

Full RAG Configuration

{
  "mcpServers": {
    "exa": {
      "command": "npx",
      "args": ["-y", "@exa/mcp-server"],
      "env": {
        "EXA_API_KEY": "your_exa_api_key"
      }
    },
    "brave-search": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
      "env": {
        "BRAVE_API_KEY": "your_brave_api_key"
      }
    },
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_TOKEN": "your_github_token"
      }
    },
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/Documents"],
      "env": {}
    },
    "postgres": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-postgres"],
      "env": {
        "POSTGRES_CONNECTION_STRING": "postgresql://localhost/knowledge_base"
      }
    }
  }
}

semantic_search_server.py

from mcp.server import Server
import numpy as np
from openai import OpenAI

server = Server("semantic-search")
client = OpenAI()

# Simulated vector store (use Pinecone, Weaviate, etc. in production)
DOCUMENTS = [
    {"id": 1, "text": "MCP enables standardized AI tool integration", "embedding": None},
    {"id": 2, "text": "Claude supports multiple MCP servers simultaneously", "embedding": None},
    # ... more documents
]

def get_embedding(text: str):
    response = client.embeddings.create(
        model="text-embedding-3-small",
        input=text
    )
    return response.data[0].embedding

def cosine_similarity(a, b):
    return np.dot(a, b) / (np.linalg.norm(a) * np.linalg.norm(b))

@server.tool()
async def semantic_search(query: str, top_k: int = 5) -> str:
    """Search documents using semantic similarity

    Args:
        query: Natural language search query
        top_k: Number of results to return (default 5)
    """
    query_embedding = get_embedding(query)

    # Compute similarities
    results = []
    for doc in DOCUMENTS:
        if doc["embedding"] is None:
            doc["embedding"] = get_embedding(doc["text"])

        similarity = cosine_similarity(query_embedding, doc["embedding"])
        results.append({"text": doc["text"], "score": similarity})

    # Sort and return top_k
    results.sort(key=lambda x: x["score"], reverse=True)
    return "\n\n".join([f"[{r['score']:.2f}] {r['text']}" for r in results[:top_k]])

if __name__ == "__main__":
    server.run()