"author:tonyweckstrom1987" matching MCP servers:

Enables AI agents to route chat completion requests to task-appropriate models via LiteLLM/OpenRouter, track usage and budgets, and run side-by-side eval comparisons of models based on cost, latency, and response quality.