Which AI providers are supported?

The skill natively supports cost and usage summaries for both Codex and Claude providers.

Does this skill track token counts?

Currently, this skill focuses on cost-only breakdowns per model, as specific token data is not split by model in the underlying CodexBar output.

What are the prerequisites for using this skill?

This skill requires the CodexBar CLI to be installed on your system (available via Homebrew Cask for macOS).

Can I use this for automated reporting?

Yes, the skill supports a --format json flag which allows you to pipe usage data into other scripts or monitoring tools.

How does it determine the 'current' model?

It analyzes the most recent daily entry in your logs and identifies the model with the highest cost breakdown or the last entry in the usage history.

Model Usage & Cost Analytics

Name: Model Usage & Cost Analytics
Author: clawdbot

byclawdbot

•

8,362

Analytics & Monitoring

Summarizes and tracks per-model cost usage from CodexBar local logs to monitor AI spend.

About

The Model Usage skill provides deep visibility into your AI spending by aggregating and summarizing local cost data from the CodexBar CLI. It allows users to query cost breakdowns for both Codex and Claude providers, supporting both 'current model' snapshots and full historical breakdowns. This skill is essential for developers and teams looking to optimize their LLM budgets, identify high-cost models, and maintain scriptable control over their AI consumption telemetry through easy JSON exports and CLI integration.

Key Features

Identify the most expensive 'current model' from recent daily logs
Generate comprehensive breakdowns of usage across all utilized models
Supports JSON output for integration into automated reporting workflows
Summarize costs for specific AI providers including Codex and Claude
Ingest data via file, stdin, or direct CodexBar CLI calls
8,362 GitHub stars

Use Cases

Quickly check the cost of the most recently used Claude model via terminal
Export a JSON summary of model costs for custom team dashboards
Audit monthly AI spend to identify which models are consuming the most budget

About

Key Features

Identify the most expensive 'current model' from recent daily logs
Generate comprehensive breakdowns of usage across all utilized models
Supports JSON output for integration into automated reporting workflows
Summarize costs for specific AI providers including Codex and Claude
Ingest data via file, stdin, or direct CodexBar CLI calls
8,362 GitHub stars

Use Cases

Quickly check the cost of the most recently used Claude model via terminal
Export a JSON summary of model costs for custom team dashboards
Audit monthly AI spend to identify which models are consuming the most budget