model-usage

Summarize per-model cost usage from CodexBar cost JSON.

Updated Apr 3, 2026
One-click install
npx skills add https://github.com/BEIRUX/agent-skills --skill model-usage-beirux
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: model-usage
Source: https://github.com/BEIRUX/agent-skills/tree/main/model-usage
Command: npx skills add https://github.com/BEIRUX/agent-skills --skill model-usage-beirux

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Provides a simple way to extract and summarize per-model usage and cost from CodexBar local cost logs so you can understand which models drive spend without manual JSON inspection.

Core Features & Use Cases

  • Current model detection: Automatically picks the most recent high-cost model or accepts an explicit model override.
  • All-model aggregation: Aggregates cost across daily rows to produce a ranked per-model cost breakdown for Codex or Claude.
  • Use Case: Generate a short report showing the current model's daily cost and a full breakdown for budgeting or engineering cost reviews.

Quick Start

Summarize per-model costs from a CodexBar cost JSON and return the current model's cost breakdown.

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 API costs from CodexBar logs?

To summarize per-model API costs from CodexBar logs, you can parse the local cost JSON via the CLI or stdin to output a ranked cost breakdown for budgeting and engineering reviews.

Can I get a cost breakdown for a specific AI model like Claude or Codex?

Yes, you can get a cost breakdown for Claude or Codex by using provider and model flags to filter the CodexBar cost JSON, returning either the most recent model summary or a full breakdown.

How do I extract AI usage costs from a JSON file for reporting?

You can extract AI usage costs from a JSON file for reporting by feeding the CodexBar cost log through the CLI, which aggregates daily rows and outputs text or machine-readable JSON.

Does this model-usage tool support filtering cost data by time range?

Yes, the model-usage tool supports filtering cost data by time range using a days flag, allowing you to limit the per-model cost aggregation to a specific recent period.

What is the best way to automate AI cost analysis from CodexBar data?

The best way to automate AI cost analysis from CodexBar data is to pipe the cost JSON via stdin into a script, using format flags to output machine-readable JSON for downstream processing.

How does current model detection work when parsing CodexBar cost logs?

Current model detection works by automatically identifying and selecting the most recent high-cost model from the CodexBar logs, unless an explicit model override is provided via CLI flags.