How do I know if the improvements worked?

The skill includes effectiveness tracking that compares error rates and success metrics from before and after the update deployment.

Can it automatically change my code?

For safe updates like adding error handling or validation, it can auto-apply changes; complex logic changes require human review.

How does this skill use telemetry?

It queries enhanced telemetry data from Loki to track tool results, errors, session patterns, and prompt usage to find improvement areas.

What tools does it integrate with?

It works seamlessly with observability-analyzer, pattern-detector, skill-updater, and skill-tester within the Skrillz ecosystem.

Does this help with token costs?

Yes, by detecting and eliminating redundant tool sequences like repeated 'Read' operations, it reduces unnecessary context usage and costs.

Observability-Driven Skill Improvement

Name: Observability-Driven Skill Improvement
Author: adaptationio

byadaptationio

0•

Analytics & Monitoring

Automates the self-improvement lifecycle of Claude skills by analyzing telemetry patterns and applying performance-enhancing updates.

This skill establishes a continuous feedback loop that bridges the gap between observability data and skill performance. By leveraging enhanced telemetry from Loki and running sophisticated pattern detection, it identifies common failure points, inefficient tool sequences, and context management issues. It then maps these patterns to actionable recommendations, allowing for automated or reviewed updates to existing skills, effectively creating a self-healing and self-optimizing environment for AI-driven development workflows.

Key Features

01Safety-first classification for auto-applying vs. manual review changes

02Intelligent pattern detection for tool failures and inefficiencies

030 GitHub stars

04Direct mapping of observability insights to specific skill updates

05Automated telemetry analysis using LogQL queries

06Performance tracking and effectiveness metrics for deployed improvements

Use Cases

01Optimizing token usage and context management by identifying and fixing redundant tool calls.

02Enhancing Bash command reliability by detecting environment-specific failures and adding prerequisite checks.

03Reducing repetitive 'file not found' errors by automatically adding existence checks to file-handling skills.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add adaptationio/skrillz skill-improvement-from-observability

For use in Claude.ai and ChatGPT

Download Skill