williamzujkowski/strudel-mcp-server

This project sits at the intersection of three established domains (MCP, Strudel live coding, LLM control APIs) and combines them in a working integration. The 184 stars and 232-day age indicate moderate adoption and proof of concept viability, but the zero velocity suggests development has stalled. The defensibility is constrained by several factors: (1) MCP itself is a new but rapidly evolving protocol that Anthropic controls—the server pattern is becoming a commodity as more integrations ship; (2) Strudel.cc is a third-party service, making this a thin orchestration layer; (3) the core capability (giving an LLM control over music generation) is not technically novel, just a new domain application; (4) switching costs are low—another developer could rebuild this in days once the MCP spec stabilizes. Frontier risk is HIGH because: Anthropic is actively promoting MCP as a platform and could trivially add Strudel integration as a first-party feature or guide, OpenAI could build similar functionality into their GPT integrations, and Claude's native abilities will likely subsume specialized music-control servers. The project has real utility for its niche (live coders + LLM enthusiasts) but lacks a defensible moat. It's a well-executed adapter rather than a novel capability.