model-usage

Name: model-usage
Rating: 0 (1 reviews)
Author: Unknown

by Unknown v1.0.0

Use CodexBar CLI local cost usage to summarize per-model usage for Codex or Claude, including the current (most recent) model or a full model breakdown. Trigger when asked for model-level usage/cost data from codexbar, or when you need a scriptable per-model summary from codexbar cost JSON.

model usage

What It Does

Analyzes CodexBar cost data to generate per-model cost summaries for Codex and Claude, either for the most recent usage or a complete breakdown.

When To Use

When you need to understand the cost breakdown by model for Codex or Claude, especially when using CodexBar for cost tracking. Use it to identify the most expensive models or to analyze overall usage patterns.

Inputs

CodexBar cost data, either directly from the CLI (using the `codexbar cost` command) or from a JSON file containing the cost data. Optional flags to specify the provider (Codex or Claude), the mode (current or all), the input file, the output format (text or JSON), and a specific model name.

Outputs

A summary of per-model cost usage, either in text or JSON format. The output includes the cost associated with each model.

Limitations

Token usage is not split by model in CodexBar output. Linux CLI support requires further documentation of the CodexBar CLI install path for Linux. Requires CodexBar to be installed and configured.

Installation

1. Ensure you have CodexBar installed and configured: `brew install steipete/tap/codexbar`. 2. Clone the OpenClaw repository containing the skill's script. 3. Navigate to the skill's directory in the cloned repository. 4. Run the script using `python {baseDir}/scripts/model_usage.py` with the appropriate arguments (e.g., `--provider codex --mode current`).

View Cline documentation

0 Installs

0 Stars

0% Success Rate

0 Trust Score

View on GitHub