A comprehensive reference implementation that demonstrates every capability of the mctx framework in one well-commented file. You get five tools covering sync returns, structured objects, progress streaming, LLM sampling via `res.ask`, and user identity through `mctx.userId`. Resources show both static URIs and dynamic templates. Prompts illustrate single-message and multi-message conversation patterns. Every tool includes annotation hints with rationale, structured logging throughout, and full test coverage via JSON-RPC requests. Clone it to study the patterns, fork it as a starting point for your own server, or reference it when you need to see how progress notifications or LLM delegation actually work in practice.
claude mcp add --transport sse mctx-ai-example-app https://example-app.mctx.ai/v1.3.4