streamable-http-mcp-server

Create and configure Streamable HTTP MCP server connections for OpenAI Agents SDK.

Updated Dec 7, 2025
One-click install
npx skills add https://github.com/syeda-hoorain-ali/todo-spec-driven-hackathon --skill streamable-http-mcp-server
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: streamable-http-mcp-server
Source: https://github.com/syeda-hoorain-ali/todo-spec-driven-hackathon/tree/main/.claude/skills/streamable-http-mcp-server
Command: npx skills add https://github.com/syeda-hoorain-ali/todo-spec-driven-hackathon --skill streamable-http-mcp-server

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Simplifies creating and configuring Streamable HTTP MCP server connections for the OpenAI Agents SDK, reducing boilerplate and setup friction.

Core Features & Use Cases

  • Create MCPServerStreamableHttp configurations with URL, headers, and timeouts, plus advanced options like caching and retry policies.
  • Manage direct HTTP MCP connections for low-latency, self-hosted deployments and testing.
  • Use with Agents to orchestrate MCP-based tool calls in production-grade workflows.

Quick Start

Create an MCPServerStreamableHttp instance with a server URL and authentication headers, then attach it to an Agent to start making MCP calls.

Frequently Asked Questions about streamable-http-mcp-server

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

FAQPage Schema
How do I configure a streamable HTTP MCP server for the OpenAI Agents SDK?

Configure a streamable HTTP MCP server by creating an MCPServerStreamableHttp instance with a server URL, authentication headers, and timeouts, then attach it to an Agent to orchestrate MCP-based tool calls in production workflows.

What options are available for customizing MCP HTTP server connections?

MCP HTTP server connections support custom timeouts, headers, caching via cache_tools_list, retry policies with max_retry_attempts and retry_backoff_seconds_base, tool filtering, structured content, and a message_handler for production-grade traffic management.

Can I use custom HTTP clients and manage connection termination with MCP servers?

Yes, MCP server configurations accept an httpx_client_factory for custom HTTP client instantiation and a terminate_on_close parameter to control connection lifecycle behavior during direct HTTP MCP server management.

What is the best way to handle authentication and timeouts for self-hosted MCP deployments?

Handle authentication and timeouts for self-hosted MCP deployments by passing custom headers and configuring timeout, sse_read_timeout, and client_session_timeout_seconds parameters directly in the MCPServerStreamableHttp constructor.

When should I use streamable HTTP MCP servers instead of other connection types?

Use streamable HTTP MCP servers for low-latency, self-hosted deployments and testing scenarios where developers need direct HTTP MCP server management with fine-grained control over caching, retry policies, and connection termination.

Does the streamable HTTP MCP server configuration support retry policies for failed tool calls?

Streamable HTTP MCP server configuration supports retry policies through max_retry_attempts and retry_backoff_seconds_base parameters, allowing automated handling of transient connection or tool call failures.