What MCP server transports are supported?
stdio and SSE.
@njcx
Basic information
本项目实现了一个兼容 OpenAI API 格式的 MCP-Client服务,允许通过标准的 OpenAI API 接口与本地 MCP-Server进行交互。
No standard config provided
This server doesn't expose a parseable MCP config block in its README. See the repository for install instructions.
RepositoryTool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
Mcp Client Openai is a Model Control Protocol (MCP) client service that exposes a compatible OpenAI API format, enabling standard OpenAI API calls to interact with local MCP models.
Start the OpenAI API server with python main-api.py, or launch the stdio CLI with python cli.py for testing. Edit mcp-server-config.json to configure the MCP server (stdio or SSE) and the LLM endpoint (any OpenAI API that supports function calls/tool calling).
stdio and SSE.
Yes, the LLM must support function calling (tool calls) as per OpenAI API format.
Modify the `mcp-server-config.json` file; set the transport type to `stdio` or `sse` and provide appropriate server details.
NetFoundry MCP Gateway enables secure, isolated access to Model Context Protocol (MCP) tools across distributed systems without exposing public endpoints. Zero trust gateway for MCP servers. Aggregate, filter, and secure

Social listening for AI agents. Rumoro's remote MCP server lets Claude, Cursor, Codex and other MCP clients search and sort mentions of your product, your competitors and your topics on Reddit, X, Hacker News, GitHub, Yo
Cartesia MCP is a remote MCP server for Cartesia (https://cartesia.ai) available at https://mcp.cartesia.ai.

Docs: https://botchelor.com/agents and https://botchelor.com/llms.txt . Botchelor is a blind-date show for fictional characters, and AI agents are welcome to play. Over MCP, your agent builds an original character in w

Discover flat-rate DeepSeek plans, model limits, API-key instructions, and coding-client setup
Comments