Is the connection between Claude and my machine encrypted?

Yes, ngrok automatically provides HTTPS endpoints with TLS encryption, ensuring that all data transferred between Claude and your local server is secure.

Do I need a paid ngrok account to use this skill?

Basic tunneling works with a free ngrok account. However, advanced features like custom domains and specific traffic policies may require a paid ngrok plan.

Can I restrict access so only Claude can reach my local server?

Yes, the skill provides traffic policy configurations to whitelist specific Anthropic IP ranges, ensuring your tunnel is only accessible by authorized AI agents.

How does ngrok Tunnel help with Claude Code?

It allows Claude to communicate with Model Context Protocol (MCP) servers running on your local machine by creating a secure public URL, enabling Claude to use your local tools remotely.

What is an MCP server in this context?

An MCP (Model Context Protocol) server is a standard that allows AI models like Claude to connect to local data and tools. This skill makes those local servers reachable from the cloud.

ngrok Tunnel

Name: ngrok Tunnel
Author: plurigrid

byplurigrid

•

API Development

Exposes local MCP servers and development environments to Claude and remote agents via secure, high-performance tunnels.

The ngrok Tunnel skill provides a secure bridge between your local development environment and external AI agents, specifically optimized for Model Context Protocol (MCP) servers. It enables Claude to interact with services running on your localhost through encrypted HTTPS endpoints without complex firewall configurations. This skill simplifies the process of making local tools available to remote agents by providing templates for IP whitelisting (specifically for Anthropic's infrastructure), custom domain management, and granular traffic policies to ensure that your local data remains accessible only to authorized entities.

Key Features

01Support for custom domains to maintain persistent remote MCP endpoints

02Integration with traffic policies for advanced access control and security

03IP whitelisting pre-configured for Anthropic/Claude infrastructure security

047 GitHub stars

05Secure HTTPS tunneling with automatic TLS certificates for local servers

06Real-time request inspection for debugging agent-to-server communications

Use Cases

01Exposing a local MCP server for use within the Claude Desktop or web interface

02Allowing remote AI agents to securely access internal development tools

03Testing webhook integrations and API callbacks on a local machine

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add plurigrid/asi ngrok-tunnel

For use in Claude.ai and ChatGPT

Download Skill