session-token-analysis

To analyze Claude Code session logs for token usage and cost optimization, you can use Python scripts to extract detailed metrics on input, output, and cache performance. The Skill identifies inefficiencies and provides actionable recommendations to reduce costs.

Cache hit rate in Claude Code measures how effectively context is reused across sessions, directly impacting token consumption. A higher cache hit rate means fewer new tokens are processed, leading to lower costs and more efficient context growth patterns.

Yes, you can review token usage metrics for multiple recent Claude Code sessions. The Skill supports analyzing several recent sessions or specific log files to track context growth, estimate costs based on model pricing, and identify optimization opportunities.

You can track context growth patterns to prevent unexpected Claude API costs by visualizing how context expands over a session. This analysis identifies inefficiencies in your token consumption and offers actionable recommendations for better cache performance.

Yes, you need Python scripts to estimate costs from Claude Code session logs. The Skill utilizes these scripts to process log files, calculate estimated costs based on model pricing, and generate detailed reports on cache performance and token usage.