maverik

A benchmarking tool for evaluating AI agents across multiple LLM models and MCP server configurations. Runs test suites against agents, tracks metrics like tool calls, costs, and duration, and provides visualization and comparison of agent performance.