Can I see which agents are used most frequently?

Yes, the 'agents' subcommand provides a full breakdown of agent usage frequency, duration, and model-specific performance metrics.

Is my analytics data shared with third parties?

No, all analytics data is stored locally on your machine in the ~/.claude/analytics/ directory and uses hashed IDs to ensure your privacy.

Does this skill support large historical data files?

Yes, it is optimized to use streaming jq queries for files over 50MB, ensuring fast performance even with extensive historical usage logs.

How does the skill calculate token costs?

The skill applies specific pricing formulas to your local statistics cache, calculating costs based on input tokens, output tokens, and cache savings.

Cross-Project Analytics

Name: Cross-Project Analytics
Author: yonatangross

byyonatangross

•

116

•

Analytics & Monitoring

Analyzes local OrchestKit usage data to provide detailed insights into agent performance, model costs, and development trends.

The Cross-Project Analytics skill provides a comprehensive monitoring dashboard for your local AI development environment. By querying data directly from the OrchestKit analytics directory, it generates deep-dive reports on agent frequency, model delegation trends, skill invocation counts, and token cost estimations. It is an essential tool for developers and teams looking to optimize their Claude Code workflows, audit API expenditure, and perform privacy-safe session replays without ever sending sensitive data to external servers.

Key Features

01Session replay functionality with tool-use timelines and precise timing data

02Performance tracking for agents, hooks, and team-based workflows

03116 GitHub stars

04Detailed breakdown of model delegation across Opus, Sonnet, and Haiku

05Accurate token cost estimation including cache savings and pricing formulas

06Privacy-focused local-only data processing with hashed project IDs

Use Cases

01Auditing monthly API expenditure and identifying optimization opportunities

02Identifying development bottlenecks by analyzing slow hooks and failure rates

03Visualizing peak usage hours and activity trends across multiple local projects

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yonatangross/orchestkit analytics

For use in Claude.ai and ChatGPT

Download Skill