mcp-server-sse

Configure Server-Sent Events MCP server connections for the OpenAI Agents SDK.

Updated Jan 15, 2026

Search Tags:#configuration #mcp #real-time #http #sse #agents-sdk #server-sent-events

Ready to run

npx skills add https://github.com/SalmanO7/Todo_Full-Stack --skill mcp-server-sse-salmano7

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: mcp-server-sse
Source: https://github.com/SalmanO7/Todo_Full-Stack/tree/main/.claude/skills/mcp-server-sse
Command: npx skills add https://github.com/SalmanO7/Todo_Full-Stack --skill mcp-server-sse-salmano7

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Configures and manages Server-Sent Events MCP server connections for the OpenAI Agents SDK, enabling streaming model context and tool results in real time.

Core Features & Use Cases

Establish MCP SSE connections with customizable URL, headers, and timeouts.
Support for tool filtering, structured content, and retry logic for robust communication.
Example use case: connect to an MCP server that streams model context and tool results to an AI agent during long-running conversations.

Quick Start

Connect to an MCP SSE endpoint by specifying the url and optional headers, then instantiate and run an agent that uses the MCP SSE server.

Frequently Asked Questions about mcp-server-sse

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

FAQPage Schema

How do I configure Server-Sent Events for an OpenAI Agents SDK workflow?▼

Server-Sent Events connections for the OpenAI Agents SDK are configured by specifying an MCP server URL, optional headers, and timeout parameters to enable streaming model context and tool results.

What parameters are available for customizing MCP SSE connections?▼

MCP SSE connections support configurable parameters including url, headers, timeout, sse_read_timeout, httpx_client_factory, cache_tools_list, name, tool_filter, use_structured_content, max_retry_attempts, and message_handler.

Can I filter which tools are exposed during real-time MCP streaming?▼

Yes, real-time MCP streaming connections support a tool_filter parameter, allowing you to restrict which tools are exposed to the AI agent during persistent, event-driven communication.

How does retry logic work for MCP SSE connections during network disruptions?▼

MCP SSE connections handle network disruptions using a configurable max_retry_attempts parameter, ensuring robust communication by automatically retrying failed streaming requests.

Does the OpenAI Agents SDK support structured content in SSE streaming?▼

Yes, the OpenAI Agents SDK supports structured content in SSE streaming through the use_structured_content parameter, ensuring event-driven MCP communication maintains data integrity.

When should I use Server-Sent Events instead of standard HTTP for MCP servers?▼

Server-Sent Events should be used for MCP servers when real-time, persistent, event-driven communication is required to stream model context and tool results during long-running AI conversations.