What problem does it solve? CodexBar's raw cost JSON is hard to interpret when you need to know which AI model is driving your spending. This Skill parses CodexBar local cost logs and produces a clear per-model cost breakdown, either for the current (most recently used) model or for all models. ## Core Features & Use Cases - Current Model Detection: Identifies the most recent model by picking the highest-cost model in the latest daily entry, with fallback to the last entry in modelsUsed. - All-Models Breakdown: Aggregates cost per model across daily rows, with optional filtering to the last N days. - Flexible Input and Output: Reads from the codexbar CLI, a JSON file, or stdin, and outputs human-readable text or structured JSON. - Use Case: You want to know how much you spent on each Codex model this week. Run the script with --provider codex --mode all --days 7 to get a ranked cost list per model. ## Quick Start Ask the agent to show your current Codex model usage cost using the model-usage skill.