Monitors and visualizes token consumption across subagents with real-time health status and threshold alerts.
The Token Usage Report skill provides critical visibility into resource utilization for complex Claude Code sessions. It generates a structured dashboard that breaks down token counts, context window percentages, and task durations on a per-subagent basis. By utilizing visual indicators for warning and critical thresholds, it enables developers to maintain session health, identify performance bottlenecks, and make data-driven decisions about when to decompose tasks or reset context to avoid reaching model limits.
Key Features
01Granular subagent-level token tracking
02Formatted summary tables for session retrospectives
03Context window utilization percentage reporting
04Task duration monitoring for performance analysis
0556 GitHub stars
06Visual health status indicators (warnings and critical alerts)
Use Cases
01Monitoring resource consumption during long-running AI orchestration sessions
02Performing health checks before initiating large-scale codebase refactors
03Analyzing subagent efficiency to optimize prompt engineering and costs
