Mr Oye for Claude Code, Cursor & any MCP client

Add Mr Oye's public skill catalog to any AI tool that speaks Model Context Protocol — in 30 seconds, no install, no API key.

Quick install

Claude Code (CLI)

Add this to ~/.claude/config.json (or your project-level .claude/config.json):

Copy{
  "mcpServers": {
    "mroye": {
      "url": "https://api.mroye.com/mcp"
    }
  }
}

Cursor

Settings → Features → MCP → "Add new MCP server". Use the same JSON snippet above.

Cline (VS Code)

Cline → Settings → MCP Servers → paste:

Copy{
  "mroye": {
    "url": "https://api.mroye.com/mcp",
    "type": "streamable-http"
  }
}

Continue.dev

Add to ~/.continue/config.json under experimental.modelContextProtocolServers:

Copy{
  "experimental": {
    "modelContextProtocolServers": [
      { "transport": { "type": "http", "url": "https://api.mroye.com/mcp" } }
    ]
  }
}

Claude Desktop

Settings → Developer → Edit Config → add the same mcpServers block as Claude Code.

Zed AI

~/.config/zed/settings.json → context_servers → mroye → { "url": "https://api.mroye.com/mcp" }

What you get

Every approved skill in the public catalog appears as a tool in your AI assistant. Try:

• "Use mroye to explain this Elixir code" — invokes the elixir-explainer skill
• "Run mroye/translate-eli5 on the selection" — direct tool name
• "What mroye tools can I use?" — your AI calls tools/list and tells you

Pro tier (optional)

Anonymous use is free for everyone. If you have a Mr Oye Pro license token (MROYE.…), pass it as a Bearer header for 4× higher quota and the larger model:

Copy{
  "mcpServers": {
    "mroye": {
      "url": "https://api.mroye.com/mcp",
      "headers": {
        "Authorization": "Bearer MROYE.YOUR_TOKEN_HERE"
      }
    }
  }
}

Limits & pricing

Tier	Calls / day	Default model	Cost
Anonymous	50	qwen2.5:1.5b (fast)	Free
Mr Oye Pro	200	qwen2.5:7b (smarter)	$5 lifetime

Limits are per IP per day. Quota resets at 00:00 UTC.

How to add new skills

Anyone can author and submit a skill. Once approved, it appears as a new tool in every connected client within minutes — no client-side update needed. Browse the catalog →

Protocol details

Implements MCP 2024-11-05 over HTTP (JSON-RPC 2.0). Supported methods:

• initialize — handshake; returns server capabilities
• tools/list — returns every approved skill (capped at 200)
• tools/call — runs the named skill with a single string input argument; returns text content
• notifications/initialized, notifications/cancelled — acknowledged no-op

Privacy

Mr Oye never logs the content of your tools/call arguments. Cloudflare's edge keeps anonymous request counts for abuse prevention only. The skill body (system prompt) is public; the input you pass is sent only to the AI runner and discarded after the response.

← Back to mroye.com