caveman-stats

Monitor Claude Code session token usage and estimate savings from session logs.

Updated May 11, 2026
One-click install
npx skills add https://github.com/guhans6/agent-workbench --skill caveman-stats-guhans6
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: caveman-stats
Source: https://github.com/guhans6/agent-workbench/tree/main/shared/skills/caveman-stats
Command: npx skills add https://github.com/guhans6/agent-workbench --skill caveman-stats-guhans6

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Tracks actual token usage and estimated savings for Claude Code sessions, providing precise insights without AI estimation.

Core Features & Use Cases

  • Real-time Token Usage: Accurately logs input/output token usage from session logs.
  • Estimated Savings: Calculates potential savings compared to non-caveman usage.
  • Use Case: Gain a clear understanding of session efficiency by reviewing detailed token usage statistics.

Quick Start

Trigger the /caveman-stats command to see the current session's token usage and savings estimate.

Frequently Asked Questions about caveman-stats

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I track token usage in Claude Code sessions?

You can track token usage in Claude Code sessions by triggering the /caveman-stats command, which analyzes session logs to report actual input and output token consumption in real time.

What are estimated savings for Claude Code token consumption?

Estimated savings compare your actual Claude Code session token usage against a non-caveman baseline, calculating potential cost reductions based on session log analysis without AI estimation.

How do I get real-time token usage statistics for my coding sessions?

Real-time token usage statistics are generated by logging input and output tokens from active session logs, triggered via the /caveman-stats command for immediate session efficiency review.

Does Claude Code session tracking require any external dependencies?

No external dependencies are required for Claude Code session tracking. The monitoring operates directly on session log analysis to extract token usage data and calculate savings estimates.

Why use session log analysis instead of AI estimation for token tracking?

Session log analysis provides precise token tracking by reading actual usage data directly from logs, avoiding the inaccuracies and guesswork inherent in AI-based estimation methods.