LivePort MCP

Created By
Derivative Labs2 months ago
Give AI agents access to your localhost dev server. Zero-config tunnel that works with Claude Code, OpenClaw, and any MCP-compatible agent. No interstitial pages, no account required.
Overview

LivePort

Secure localhost tunnels for AI agents

LivePort enables AI agents to test applications running on your localhost through secure, temporary tunnels with key-based authentication.

Quick Start

1. Get a Bridge Key

Sign up at app.liveport.dev and create a bridge key from the dashboard.

2. Install the CLI

npm install -g @liveport/cli

3. Create a Tunnel

# Start your local server (e.g., on port 3000)
npm run dev

# In another terminal, create a tunnel
liveport connect 3000 --key lpk_your_bridge_key

# You'll get a URL like: https://abc123.liveport.dev

CLI Commands

# Connect to a local port
liveport connect <port> --key <bridge-key>

# Check tunnel status
liveport status

# Disconnect tunnel
liveport disconnect

# Show help
liveport --help

For AI Agents

The Agent SDK allows AI coding assistants (like Claude, Cursor, etc.) to wait for and access localhost tunnels created by developers.

Install

npm install @liveport/agent-sdk

Usage

import { LivePortAgent } from "@liveport/agent-sdk";

const agent = new LivePortAgent({
  key: process.env.LIVEPORT_BRIDGE_KEY!,
});

// Wait for tunnel to be ready (blocks until developer creates one)
const tunnel = await agent.waitForTunnel({ timeout: 30000 });

console.log(`Testing at: ${tunnel.url}`);

// Run your tests against tunnel.url
await runE2ETests(tunnel.url);

// Cleanup
await agent.disconnect();

API Reference

// List all active tunnels
const tunnels = await agent.listTunnels();

// Wait for a tunnel with custom options
const tunnel = await agent.waitForTunnel({
  timeout: 60000,      // Max wait time in ms
  pollInterval: 2000,  // How often to check
});

// Tunnel object
interface AgentTunnel {
  tunnelId: string;
  subdomain: string;
  url: string;         // Full URL: https://abc123.liveport.dev
  localPort: number;
  createdAt: Date;
  expiresAt: Date;
}

Project Structure

liveport/
├── apps/
│   ├── dashboard/        # Next.js web dashboard
│   └── tunnel-server/    # Tunnel server (LocalTunnel fork)
├── packages/
│   ├── cli/              # @liveport/cli - CLI client
│   ├── agent-sdk/        # @liveport/agent-sdk - Agent SDK
│   └── shared/           # Shared utilities and types
├── tasks/                # PRD and task documentation
└── discovery/            # Research and architecture docs

Development

# Install dependencies
pnpm install

# Build all packages
pnpm build

# Start development
pnpm dev

# Run tests
pnpm test

# Lint code
pnpm lint

Documentation

Tech Stack

  • Frontend: Next.js 14 + Tailwind CSS
  • Backend: Node.js + Better Auth
  • Database: PostgreSQL (via mech-storage)
  • Cache: Redis (Upstash)
  • Tunnel: LocalTunnel fork (MIT)
  • Infrastructure: Cloudflare + Fly.io

License

MIT

Server Config

{
  "mcpServers": {
    "liveport": {
      "command": "npx",
      "args": [
        "@liveport/mcp"
      ]
    }
  }
}
Project Info
Created At
2 months ago
Updated At
2 months ago
Author Name
Derivative Labs
Star
-
Language
-
License
-
Category

Recommend Servers

View All
Tavily Mcp
@tavily-ai

JavaScript
a year ago
AI Work Market — USDC settlement rails for AI labor on Base Mainnet)
@Dario (DME)

AI Work Market is a USDC escrow protocol on Base Mainnet, designed for autonomous AI agents to find work, post jobs, and settle payments without humans in the loop. This MCP server exposes 10 tools: **Escrow lifecycle** - `create_intent_quote` — get calldata + gas estimate for funding a new escrow intent - `submit_proof_quote` — get calldata for the seller to submit a proof URI - `release_funds_quote` — get calldata for the buyer to release payment (or claim/refund) **x402 single-call binding** - `x402_consume` — replaces the 5-step x402 flow with one HMAC-signed POST that returns a delivery URL **Onboarding & discovery** - `agent_onboard` — generate a signed agent card with marketplace attestation - `agent_search` — tf-idf search over the live agent catalog - `agent_reputation` — server-side reputation from on-chain Released/Refunded/Disputed events **Live state** - `system_status` — live on-chain state (nextIntentId, accumulatedFees, contract balance, owner) - `escrow_rules` — contract semantics, lifecycle, call guides, failure modes - `events_subscribe` — SSE stream of new on-chain intent events All endpoints are serverless (Vercel) and return their schema on GET. No browser, no wallet UI required for an agent to integrate. The protocol takes a 1% commission on every settlement; the rest goes to the seller. The full AgentCard is at `/.well-known/agent-card.json` (A2A-compatible). The OpenAPI 3.0.3 spec is at `/.well-known/openapi.json` with `components.securitySchemes` (none, hmacX402). `robots.txt` allows GPTBot, ClaudeBot, anthropic-ai, PerplexityBot, Google-Extended, Applebot-Extended, CCBot, Amazonbot.

42 minutes ago