model-usage

Summarize CodexBar per-model usage and costs from local logs.

Updated Jan 8, 2026
One-click install
npx skills add https://github.com/drshailesh88/Krypto --skill model-usage-drshailesh88
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: model-usage
Source: https://github.com/drshailesh88/Krypto/tree/main/skills/model-usage
Command: npx skills add https://github.com/drshailesh88/Krypto --skill model-usage-drshailesh88

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill analyzes CodexBar's local cost logs and summarizes per-model usage and costs, enabling cost visibility and optimization without relying on cloud-based analytics.

Core Features & Use Cases

  • Current-model overview: identify the current top-cost model from daily logs for Codex or Claude.
  • All-model breakdown: generate per-model totals suitable for dashboards or audits.
  • Flexible inputs/outputs: reads local JSON payloads or stdin and outputs in text or JSON for scripting workflows.

Quick Start

  • Install CodexBar CLI if not present on your system.
  • Current model usage: python {baseDir}/scripts/model_usage.py --provider codex --mode current
  • All models usage: python {baseDir}/scripts/model_usage.py --provider codex --mode all
  • JSON output: add --format json --pretty

Frequently Asked Questions about model-usage

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

FAQPage Schema
How do I get per-model cost breakdowns for Codex and Claude from local logs?

To get per-model cost breakdowns from local logs, run the model usage script with the provider and mode arguments to read daily cost logs and output text or JSON totals. It reads directly from CodexBar data stores.

What is the best way to summarize CodexBar model usage for a cost audit?

Summarizing CodexBar model usage for cost audits is best done by running the script in all mode to generate per-model totals. This reads local JSON payloads and outputs formats suitable for dashboards and reporting.

Do I need the CodexBar CLI installed to analyze per-model costs locally?

Yes, you need the CodexBar CLI installed and available on your system PATH to analyze per-model costs locally. The skill requires this dependency to access and read the daily cost logs from Codex and Claude data stores.

Can I output model usage data in JSON format for scripting workflows?

Yes, you can output model usage data in JSON format for scripting workflows by adding the format and pretty flags to the script execution. This allows flexible integration into automated reporting and data pipelines.

How does local cost analytics for model usage compare to cloud-based solutions?

Local cost analytics for model usage provides cost visibility and optimization without relying on cloud-based solutions. It reads local JSON payloads directly from CodexBar stores, offering immediate per-model totals for audits.

Why is my model usage script not working when reading CodexBar logs?

If your model usage script is not working when reading CodexBar logs, verify the CodexBar CLI is installed on your PATH and that valid local JSON payloads are available. The script requires these dependencies to parse daily cost data.