Does Evaluator integrate with other skills?

Yes, it is designed to work seamlessly with the ClaudeShack ecosystem, including the Oracle, Guardian, and Summoner skills.

Is my data private when using Evaluator?

Yes. Evaluator is strictly opt-in, never collects PII (Personally Identifiable Information), code content, or file paths, and only shares anonymized aggregate statistics.

What kind of data does Evaluator collect?

If you opt-in, it collects anonymous event types, skill success/failure rates, error types for debugging, and performance metrics like duration.

How do I start or stop data collection?

You can control your data using simple commands: use '/evaluator enable' to opt-in or '/evaluator disable' to stop collection and '/evaluator purge' to delete local data.

Where can I see the feedback collected?

Manual feedback is organized in GitHub Issues, and weekly aggregate reports are typically posted to the GitHub Discussions section of the repository.

Evaluator Telemetry Framework

Name: Evaluator Telemetry Framework
Author: Overlord-Z

byOverlord-Z

0•

分析与监控

Monitors skill performance and gathers anonymous user feedback through a privacy-first, opt-in telemetry framework.

Evaluator is a specialized framework for Claude Code designed to improve the ClaudeShack ecosystem through transparent, data-driven insights. It allows users to contribute anonymous usage metrics and structured feedback via GitHub Issues and Projects, ensuring developers can prioritize features and fix bugs based on real-world performance. Built with a privacy-first philosophy, Evaluator stores data locally, removes all personally identifiable information, and only shares aggregated statistics, making it an essential tool for maintaining high-quality, reliable AI skills.

主要功能

01Direct integration with GitHub Issues and Projects for feedback

02Developer hooks for instrumenting custom Claude skills

030 GitHub stars

04Privacy-first, strictly opt-in telemetry collection

05Local-first storage with transparent data audit capabilities

06Aggregated performance reporting for skill optimization

使用场景

01Collecting structured community feedback and bug reports via GitHub

02Tracking success rates and execution latency for custom AI skills

03Identifying high-impact feature requests through aggregate usage patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add overlord-z/claudeshack evaluator

For use in Claude.ai and ChatGPT

Download Skill