MCP of MCPs

MCP of MCPs is a meta-server that merges all your MCP servers into a single smart endpoint. It gives AI agents instant tool discovery, selective schema loading, and massively cheaper execution, so you stop wasting tokens and time.

With persistent tool metadata, semantic search, and direct code execution between tools, it turns chaotic multi-server setups into a fast, efficient, hallucination-free workflow. It also automatically analyzes the tools output schemas if needed and preserves them across sessions for consistent behavior.

In short: 🚀 Faster automation 🧠 Cleaner reasoning 💰 Drastically fewer tokens 📦 Persistent + analyzed schema metadata

Tool 1: semantic_search_tools

Semantic Discovery Tool - Search tools by describing the task you want to accomplish. Instead of browsing all of tool names—or when tool names don’t clearly indicate what they do—just describe your intent in plain English (e.g., “send notifications”, “query database”, “process images”) and get back only the most relevant tools instantly. This provides a fast and lightweight approach to investigate what tools are available across all connected servers without loading any full tool definitions.

// Search by task/intent, not by tool names:
// Input: { query: "send notifications to a channel", limit: 5 }
// Returns only relevant matches (ranked by similarity):
// [
//   {
//     serverName: "slack",
//     toolName: "post_message",
//     description: "Post a message to a Slack channel",
//     similarityScore: 0.94,
//     fullPath: "slack/post_message"
//   },
// ....
//   // Only 5 most relevant tools returned - fast and lightweight!
// ]

Perfect for quick investigation:

• Describe what you need to do, not what tool you need
• Get instant results without loading full schemas
• Discover capabilities across all servers in milliseconds
• No token overhead - just lightweight tool names and descriptions

Tool 2: get_mcps_servers_overview

Discovery Tool - This tool returns only tool names without full schemas, giving agents a lightweight overview in seconds instead of loading hundreds of detailed definitions upfront. By showing just what’s available without overwhelming details, it prevents confusion and hallucinations while eliminating loading delays.

// Returns:
// google_drive/download_file
// google_drive/upload_file
// slack/send_message
// database/execute_query
// ...

Tool 3: get_tools_overview

Introspection Tool - Load only the tools you actually need instead of all 30+ tools, saving thousands of tokens through selective loading. This on-demand approach provides faster responses and focused context that reduces confusion and improves accuracy.

// Input: ["google_drive/download_file", "slack/send_message"]
// Returns: Full tool definitions with:
// - Parameter schemas
// - Required vs optional fields