How does Pattern Learning improve Claude's performance?

It records the success rate and quality scores of completed tasks, allowing Claude to reuse proven strategies and automatically load the most effective skills for similar tasks in the future.

Where is the pattern data stored?

All pattern data, task history, and effectiveness metrics are stored locally in your project directory under the '.claude-patterns/' folder, ensuring total privacy.

Can I see the metrics for different skills?

Yes, the skill maintains a 'skill-effectiveness.json' file that provides transparency into how often specific skills were used and their contribution to successful outcomes.

Does this work with any programming language?

Yes, Pattern Learning is language-agnostic. It automatically detects the languages and frameworks in your repository to categorize patterns accurately.

Autonomous Pattern Learning

Name: Autonomous Pattern Learning
Author: bejranonda

bybejranonda

•

生産性とワークフロー

Automates project-specific pattern recognition and retrieval to optimize Claude's execution strategy through continuous self-learning.

Pattern Learning is an intelligent framework for Claude Code that builds a persistent project knowledge base by recording and analyzing successful task executions. It automatically categorizes development activities—such as refactoring, bug fixes, or feature implementation—and tracks the specific context, tools used, and resulting quality metrics. By maintaining a 100% local database of what works best for your specific codebase, it enables Claude to intelligently recommend the most effective skills and strategies for future tasks, ensuring continuous improvement and consistent adherence to your project's unique architectural patterns.

主な機能

01Persistent local pattern database stored within your project directory

02Comprehensive quality assessment scoring including test passing and standards compliance

03Automatic task categorization and context extraction (languages, frameworks, structure)

049 GitHub stars

05Privacy-first, 100% local processing of project-specific insights

06Dynamic skill auto-selection algorithm based on historical success rates

ユースケース

01Automatically selecting the best refactoring tools for a legacy codebase based on previous successes

02Ensuring consistent architectural pattern adherence across large-scale microservices

03Optimizing bug-fix strategies by analyzing historical resolution patterns in specific modules

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bejranonda/llm-autonomous-agent-plugin-for-claude pattern-learning

For use in Claude.ai and ChatGPT

Download Skill