I Built a Local Usage Intelligence Dashboard for Claude Code

Описание и применение

Разработчики получили локальный дашборд для анализа токенов и стоимости использования Claude Code без передачи данных в облако.

Задача

Developers need to understand token consumption, cost, and cache behavior of Claude Code without sending data to external services.

Результат

Engineers can now see token consumption per project, identify expensive sessions, verify cache effectiveness, and track usage trends—all locally and privately.

Рабочий процесс

• 1. Install CC Token Meter from GitHub.
• 2. Run the tool to parse Claude Code’s local transcript files.
• 3. The parser builds a private usage index.
• 4. Analytics compute token counts, estimated cost, cache hits, and project attribution.
• 5. A local web dashboard visualizes the insights.
• 6. Review dashboards to identify high‑cost projects, sessions, and cache inefficiencies.

Доказательства

Instead of intercepting Claude Code requests or introducing another cloud service, CC Token Meter works from the session data Claude Code already stores locally.

Источник