hosted-mcp-server

Configure hosted MCP server connections for the OpenAI Agents SDK.

Updated Jan 15, 2026
One-click install
npx skills add https://github.com/SalmanO7/Todo_Full-Stack --skill hosted-mcp-server-salmano7
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hosted-mcp-server
Source: https://github.com/SalmanO7/Todo_Full-Stack/tree/main/.claude/skills/hosted-mcp-server
Command: npx skills add https://github.com/SalmanO7/Todo_Full-Stack --skill hosted-mcp-server-salmano7

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill simplifies connecting and configuring hosted Model Context Protocol (MCP) server connections for the OpenAI Agents SDK, enabling secure, auditable tool execution.

Core Features & Use Cases

  • Create HostedMCPTool configurations with server_label and server_url
  • Manage server authentication, connectors, and optional headers for MCP servers
  • Define tool approval policies to control sensitive actions
  • Supports connector-backed tools and public MCP servers

Quick Start

Provide a server_label and server_url to establish a hosted MCP connection for your agent workflow.

Frequently Asked Questions about hosted-mcp-server

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I configure a hosted MCP server connection for OpenAI Agents?

To configure a hosted MCP server, provide a server_label and server_url to establish a secure connection for your OpenAI Agents SDK workflow. This creates an auditable HostedMCPTool configuration for tool execution.

What is a hosted MCP server and when do I need it for AI agents?

A hosted MCP server provides secure, auditable Model Context Protocol integration for AI agents. You need it when your OpenAI Agents require external tool execution with strict authentication and approval flow management.

Can I use connector-backed tools and custom headers with MCP servers?

Yes, you can use connector-backed tools and optional headers with public or connector-backed MCP servers. The configuration manages server authentication and connectors to enable secure tool execution within your agent workflow.

How do I define tool approval policies for sensitive actions in OpenAI Agents?

You define tool approval policies during the hosted MCP server configuration process. This controls sensitive actions by applying approval flow requirements to the connected HostedMCPTool, ensuring auditable tool execution.

Does the OpenAI Agents SDK require hosted MCP connections for secure tool execution?

The OpenAI Agents SDK uses hosted MCP connections to ensure secure and auditable tool execution. Applying this configuration manages server authentication and approval flows required for sensitive agent actions.