hello all ive been working on little MCP server so that LLM like Claude or even localized olama model can control and do some of your audio processing for you cutting your workload down. would be amazing to get feedback on it so i can improve it.