home / mcp / hypertool mcp server

HyperTool MCP Server

Dynamically expose tools from proxied servers based on an Agent Persona

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "toolprint-hypertool-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@toolprint/hypertool-mcp",
        "mcp",
        "run",
        "--mcp-config",
        ".mcp.hypertool.json"
      ]
    }
  }
}

HyperTool MCP lets you orchestrate tools from many MCP servers so your AI can work with the exact toolsets you need. It unlimits the total number of tools you can connect, enables task-focused toolsets, and adds smart tool descriptions to help your AI pick the right tool for the job.

How to use

Use HyperTool MCP to connect multiple MCP servers and switch between focused toolsets. Youโ€™ll equip a toolset for a specific task and your AI will operate with the most relevant tools from your connected MCP servers.

How to install

Prerequisites: you need Node.js 18+ and npm (or a compatible package manager) installed on your machine.

1) Clone or install HyperTool MCP from its package sources and install dependencies.

2) Run the setup steps to get dependencies and pre-commit hooks in place.

3) Use the provided MCP command to start or connect to the HyperTool MCP server as described in the configuration steps.

Additional configuration and usage notes

- Dynamic Toolsets: Build, modify, and switch between toolsets on the fly to match your current task.

- Personas: Use pre-configured bundles that include MCP servers and curated toolsets for common workflows like web development, data science, and DevOps.

- Context measurement: See token estimates for each tool to optimize your toolsets and minimize context usage.

- Configuration Mode: Keep toolset management separate from operational tools to maintain focus and reduce confusion.

- HTTP mode: Run HyperTool as a long-lived HTTP server for persistent connections and faster responses.

Real-World Toolsets

Create focused toolsets for different workflows, such as development, content creation, or research.

Examples include naming toolsets like "deep-coding" (git, docker, filesystem) or "writing" (Notion, grammarly, slack).

App compatibility and hot-swap behavior

HyperTool works with MCP-compatible apps. You can switch toolsets on the fly in supported apps like Cursor/VSCode without restarting, while some clients may require a restart for certain toolsets.

All features

Explore the full range of capabilities, including personas, server groups, context measurement, configuration mode, dynamic toolsets, tool annotations, HTTP mode, and unlimited servers.

Troubleshooting and tips

If a server goes down, HyperTool monitors health and reconnects when itโ€™s back. Toolsets stay intact and you can continue working once connectivity is restored.

Development and contribution

The project uses Node.js tooling and offers commands to build, test, lint, format, and run type checks. Pre-commit hooks help maintain code quality.

Security and data handling

Token usage and context information are surfaced to help you optimize tool usage. Token estimates are approximations and are cached for performance.

Available tools

git

Version control operations used within toolsets, including status, commit, push, and related workflows.

docker

Containerization tools for building, running, and managing containers across toolsets.

filesystem

Access to local filesystem for reading and writing files, scripts, and artifacts.

browser

Web browser automation or interactions to fetch data and tests.

testing

Tools for testing code and deployments.

notion

Note-taking and collaboration integration for creating and updating documents.

grammarly

Grammar and writing quality checks to improve text output.

slack

Team communication integration for notifications and messaging.

perplexity

AI-assisted information access and question answering.

arxiv

Academic paper search and retrieval via ArXiv integration.

wikipedia

Encyclopedic information access via Wikipedia.

traces

Tracing and debugging tools to monitor execution paths.

curl

Command-line data transfer utility used for API checks and data fetches.

templates

Pre-built document or message templates for rapid content creation.