MCP Essentials Setup

Three MCP servers that enhance Claude Code workflows.

Quick Install

npm i -g mcp-omnisearch mcp-sqlite-tools mcpick

1. mcp-omnisearch

Unified search across Tavily, Brave, Kagi, Perplexity, GitHub, and more.

GitHub: https://github.com/spences10/mcp-omnisearch

Config (~/.claude/settings.json):

{
  "mcpServers": {
    "mcp-omnisearch": {
      "command": "npx",
      "args": ["-y", "mcp-omnisearch"],
      "env": {
        "TAVILY_API_KEY": "your-key",
        "BRAVE_API_KEY": "your-key",
        "KAGI_API_KEY": "your-key"
      }
    }
  }
}

Use when: Web search, GitHub code search, AI-powered answers, content extraction.

2. mcp-sqlite-tools

Safe SQLite operations with read/write separation and transaction support.

GitHub: https://github.com/spences10/mcp-sqlite-tools

Config:

{
  "mcpServers": {
    "mcp-sqlite-tools": {
      "command": "npx",
      "args": ["-y", "mcp-sqlite-tools"]
    }
  }
}

Use when: Query databases, analyze data, manage SQLite files.

3. mcpick

Dynamically enable/disable MCP servers to optimize context usage.

GitHub: https://github.com/spences10/mcpick

Install: npm i -g mcpick

Usage:

mcpick enable omnisearch sqlite-tools
mcpick disable memory-server
mcpick list

Use when: Too many MCPs eating context, need to toggle servers per-project.

Recommended Combinations

Workflow	Enable
Research	omnisearch
Data analysis	sqlite-tools
Full stack	omnisearch + sqlite-tools
Minimal context	Use mcpick to toggle as needed