tokencast

tokencast is an MCP server designed for AI coding environments that lets users estimate the token and dollar cost of planned large language model (LLM) agent workflows before actual execution. It integrates with Claude Code, Cursor, VS Code (with GitHub Copilot), and Windsurf, exposing tools for pre-task cost estimation, calibration tracking, and post-execution cost reporting. This makes it ideal for developers and AI workflow engineers who want budget-aware planning and continuous calibration of LLM agent usage in their IDEs.