model-usage

Summarize per-model CodexBar cost data from local logs.

1|Updated Jan 28, 2026
One-click install
npx skills add https://github.com/Zentin-L/Masterbot --skill model-usage-zentin-l
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: model-usage
Source: https://github.com/Zentin-L/Masterbot/tree/main/skills/model-usage
Command: npx skills add https://github.com/Zentin-L/Masterbot --skill model-usage-zentin-l

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

CodexBar cost data is generated locally; this skill provides a concise, per-model cost summary to help optimize spending and compare models.

Core Features & Use Cases

  • Current-model snapshot: quickly identify the most expensive active model for a provider (Codex or Claude) using the latest daily entry.
  • All-model breakdown: compute total costs by model across all available daily rows to inform cost-aware decisions.
  • Flexible inputs: accepts direct CodexBar JSON payloads, files, or reads from the CodexBar CLI on PATH.

Quick Start

Run the model_usage.py script with appropriate flags to generate a per-model cost report for a chosen provider.

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 CodexBar costs from local logs?

You can break down per-model CodexBar costs by applying the model_usage script to local JSON payloads or CLI data, producing either human-readable text or JSON spending summaries.

How do I find the most expensive active model in my CodexBar cost data?

You can identify the most expensive active model by generating a current-model snapshot from the latest daily CodexBar entry to reveal the top spending model for a chosen provider.

Does the CodexBar cost summarizer require the CLI to be installed?

The CodexBar cost summarizer supports direct JSON payloads or files as input, but can also read from a locally installed CodexBar CLI if it is available on your PATH.

Can I limit the per-model cost breakdown to just the last few days?

You can limit the per-model cost breakdown to recent days by applying an optional day limiting parameter when running the model_usage script against your local CodexBar logs.

What is the best way to compare total spending across all available AI models?

To compare total spending across all models, compute an all-model breakdown that aggregates total costs by model across all available daily rows in your local CodexBar logs.