RaptorWayz MCP server

An MCP server that lets any MCP-compatible AI agent get past Cloudflare challenges. It exposes one tool:

solve_cloudflare(domain) → returns a valid cf_clearance cookie, the matching User-Agent, and a proxy on the same exit IP. The agent attaches all three to its own request to load the protected page. RaptorWayz never sees your traffic (the proxy does no TLS interception). This is cookie+proxy, not fetch-on-behalf.

Setup

You need a RaptorWayz API key. Get one (5 free solves) at https://raptorwayz.com.

Claude Desktop / Claude Code / any MCP client

Add to your MCP config (e.g. claude_desktop_config.json):

{
  "mcpServers": {
    "raptorwayz": {
      "command": "npx",
      "args": ["-y", "raptorwayz-mcp"],
      "env": {
        "RAPTORWAYZ_API_KEY": "bk_your_key_here"
      }
    }
  }
}

Or, if installed globally (npm i -g raptorwayz-mcp), use "command": "raptorwayz-mcp".

Environment variables

| Var | Required | Default | Notes | |-----|----------|---------|-------| | RAPTORWAYZ_API_KEY | yes | (none) | Your API key (Bearer credential). | | RAPTORWAYZ_API_BASE | no | https://raptorwayz.com | Override the API base URL. |

Using the tool

Ask your agent something like "fetch https://example.com but it's behind Cloudflare". The agent calls solve_cloudflare("example.com") and gets back the clearance triple plus a ready-to-run snippet:

import requests
resp = requests.get(
    "https://example.com/",
    headers={"User-Agent": "<returned user_agent>"},
    cookies={"cf_clearance": "<returned cf_clearance>"},
    proxies={"https": "<returned proxy>"},
)

All three (cookie + User-Agent + proxy) must be used together or Cloudflare re-challenges.

Develop

npm install
npm run build
RAPTORWAYZ_API_KEY=bk_... node dist/index.js   # stdio server

Registry listings

Manifests for submission are in this folder: server.json (official MCP registry) and smithery.yaml (Smithery). Validate against each registry's latest schema before submitting.

MIT licensed.