"author:karmendra8386" matching MCP servers:

Enables agents and developers to estimate, budget, and compare LLM spend per task before and during a run, including ranking models by expected cost per successful task and tracking spend in a local JSON ledger.