# Build authenticated MCP tools and servers with Arcade MCP

> Use Arcade MCP to create custom MCP servers and tools with OAuth-aware authorization, evals, and deployment paths for agent tool-calling workflows.

- Skill: `agentskillexchange/build-authenticated-mcp-tools-and-servers-with-arcade-mcp` (Agent Skill)
- Install (CLI): `npx skillmds@latest add agentskillexchange/build-authenticated-mcp-tools-and-servers-with-arcade-mcp`
- Raw SKILL.md: https://api.skillmd.com/api/skills/agentskillexchange/build-authenticated-mcp-tools-and-servers-with-arcade-mcp/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: agentskillexchange (https://skillmd.com/u/agentskillexchange)
- Updated: 2026-09-08
- Page: https://skillmd.com/skills/agentskillexchange/build-authenticated-mcp-tools-and-servers-with-arcade-mcp

---


# Build authenticated MCP tools and servers with Arcade MCP

Use Arcade MCP to create custom MCP servers and tools with OAuth-aware authorization, evals, and deployment paths for agent tool-calling workflows.

## Prerequisites

Python, arcade-mcp, an MCP-compatible client, optional Arcade Cloud account for hosted auth and deployment

## Installation

Use the upstream install or setup path that matches your environment:
- uv tool install arcade-mcp
- uv run server.py # stdio (default), for Claude Desktop and CLI tools
- uv run server.py http # HTTP+SSE, for Cursor and VS Code; docs at http://127.0.0.1:8000/docs
- git clone https://github.com/ArcadeAI/arcade-mcp.git

Requirements and caveats from upstream:
- [![Python Version](https://img.shields.io/pypi/pyversions/arcade-mcp)](https://pypi.org/project/arcade-mcp/)
- **Open-source Python framework for building MCP servers and tools.**
- arcade-mcp is the Python framework for building [Model Context Protocol](https://modelcontextprotocol.io) servers and the tools that run inside them. It powers the 7,500+ prebuilt tools across 81 MCP servers at [Arcad...

Basic usage or getting-started notes:
- The secrets (OAuth tokens, API keys, etc) are securely injected by Arcade into your tool call at runtime, so that your tool can authorize requests to upstream APIs, for example. **The client and the LLM never see the...
- bash
- ### Scaffold a new server

- Source: https://github.com/ArcadeAI/arcade-mcp
- Extracted from upstream docs: https://raw.githubusercontent.com/ArcadeAI/arcade-mcp/HEAD/README.md

## Documentation

- https://docs.arcade.dev

## Source

- [Agent Skill Exchange](https://agentskillexchange.com/skills/build-authenticated-mcp-tools-and-servers-with-arcade-mcp/)