skills.rest
ferrarifankid04
ferrarifankid04COMMUNITY
Explore all skills by this author →
GitHub

caveman-stats

Displays real token usage and estimated savings from the Claude Code session log.

Updated Aug 6, 2026
Search Tags:#claude code#statistics#token usage#cost tracking#hook#session log
One-click install
npx skills add https://github.com/ferrarifankid04/ai-skills-public --skill caveman-stats-ferrarifankid04
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman-stats
Source: https://github.com/ferrarifankid04/ai-skills-public/tree/main/antigravity/caveman/plugins/caveman/skills/caveman-stats
Command: npx skills add https://github.com/ferrarifankid04/ai-skills-public --skill caveman-stats-ferrarifankid04

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Users running token-saving modes in Claude Code have no visibility into how many tokens they actually consumed or saved during a session, since estimates from the model itself are unreliable. ## Core Features & Use Cases - Real Usage Reporting: Reads token counts directly from the Claude Code session log instead of relying on AI estimation. - Savings Estimation: Shows estimated token savings for the current session alongside actual usage. - Hook-Driven Output: Triggered by the /caveman-stats command, with output injected by the mode-tracker hook so the model performs no computation. - Use Case: After a long coding session with a compressed-output mode enabled, run /caveman-stats to see exactly how many tokens were used and how many were saved. ## Quick Start Ask the assistant to run /caveman-stats to show the current session's token usage and estimated savings.

Frequently Asked Questions about caveman-stats

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

FAQPage Schema
How do I check token usage in a Claude Code session?▼

Run the /caveman-stats command during your session. The mode-tracker hook reads the Claude Code session log and returns the real token usage plus estimated savings immediately, without any AI estimation.

How are token savings calculated in caveman-stats?▼

Savings are computed from actual token counts read directly from the Claude Code session log, not from model guesses. The caveman-stats hook formats these numbers and returns them as the command output.

Does caveman-stats require the model to compute anything?▼

No. The hook returns decision: "block" with the formatted stats as the reason, so the model does nothing when the skill fires. The user sees the numbers immediately from the hook output.

Why are my caveman-stats numbers different from what the model estimated?▼

Model estimates of token usage are unreliable because the model cannot see its own billing counts. caveman-stats reads the authoritative session log instead, so its figures reflect actual recorded usage.

Related Skills

Explore More Software Engineering Skills→
Bonny94ITABonny94ITA

caveman-stats

Analyze Claude Code session logs to report token usage and savings.

Community
Basic
tombeliebertombelieber
101

daily-cost

Monitor daily Claude Code spending with aggregate and per-session cost breakdowns.

Community
Intermediate
ttphatsttphats

token-budget-advisor

Estimates prompt token counts and offers selectable response depth levels before answering.

Community
Intermediate
Shana-AEShana-AE
2

qiniu-usage

Query Qiniu AI API token usage and compute USD/CNY costs.

Community
Advanced