What problem does it solve? Codex usage data is scattered across raw session transcripts, making it hard to see where tokens, cache, and subagent costs actually go. This Skill turns those transcripts into a single explorable HTML report with totals, breakdowns, and anomaly callouts. ## Core Features & Use Cases - Usage Aggregation: Runs a bundled Node.js analyzer over ~/.Codex/projects transcripts to compute token totals, per-project and per-skill breakdowns, cache-hit rates, and top prompts. - Interactive HTML Report: Embeds the JSON into a self-contained template with sortable tables, drill-downs, and block-char bar charts. - Anomaly Narrative: Fills dedicated sections with findings such as disproportionate project spend, low cache-hit rates, or oversized prompts, plus optimization suggestions. - Use Case: After a heavy week of Codex work, generate a report to discover that one project consumed 41% of tokens across three sessions and that cache-hit rate dropped below 85%. ## Quick Start Generate a session report of my Codex usage for the last 7 days and save the HTML file here.
