model-usage

Extract per-model cost data from CodexBar local cost logs.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/kulharir7/karya --skill model-usage-kulharir7
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: model-usage
Source: https://github.com/kulharir7/karya/tree/main/workspace/plugins/model-usage
Command: npx skills add https://github.com/kulharir7/karya --skill model-usage-kulharir7

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

CodexBar's local cost logs can be difficult to parse manually. This Skill automatically extracts per-model usage data to help you understand cost distribution across models.

Core Features & Use Cases

  • Per-model cost summaries: Get current-model and all-model breakdowns for Codex and Claude.
  • Flexible output: Produce text summaries or JSON for downstream processing.
  • Use Case: Quickly review which model consumed the most cost in a given period to guide optimization or budgeting decisions.

Quick Start

Run the bundled script to summarize per-model costs from CodexBar's local cost JSON (current model or all models).

Frequently Asked Questions about model-usage

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

FAQPage Schema
How do I summarize per-model cost data from CodexBar logs?

To summarize per-model cost data from CodexBar logs, this Skill reads the local JSON cost file and extracts cost breakdowns for Codex and Claude. It processes the most recent daily entry to report the current model or aggregates all models based on your requested view.

What is the best way to parse local CodexBar JSON usage files?

Parsing local CodexBar JSON usage files is best handled by running the bundled script, which automatically extracts cost distribution data and renders a concise summary in text or JSON format for downstream processing.

Can I get a JSON breakdown of Codex and Claude model costs instead of text?

Yes, you can get a JSON breakdown of Codex and Claude model costs. The Skill supports flexible output formats, allowing you to produce either a concise text summary or structured JSON data depending on your downstream processing needs.

Does this per-model usage tool require external dependencies to analyze costs?

No, analyzing per-model usage does not require external dependencies. The Skill operates entirely on CodexBar's local cost logs, reading the JSON cost file directly to extract and aggregate the cost data for your models.

When should I use a local cost log analysis instead of manual review?

Local cost log analysis is needed when CodexBar's raw JSON files become difficult to parse manually. Use it to quickly review which model consumed the most cost in a given period to guide optimization or budgeting decisions.

How do I check which AI model consumed the most cost in my daily logs?

To check which AI model consumed the most cost in your daily logs, the Skill analyzes the most recent daily entry in CodexBar's local cost JSON. It aggregates costs across all models for Codex and Claude to identify the highest spender.