How does Knowledge Extractor persist data?

It uses a Kuzu-based memory backend to store discoveries and updates local markdown files like PATTERNS.md and DISCOVERIES.md to ensure information remains accessible to both humans and AI.

When is the best time to use this skill?

It is most effective at the end of a session, immediately after solving a complex bug, or when you have successfully implemented a feature that follows a clear, reusable logic.

Does this skill help with agentic workflow automation?

Yes, it specifically monitors for workflows you have repeated 2-3 times and provides a structured blueprint for creating a new specialized agent to handle those tasks.

Can I customize the templates used for knowledge extraction?

The skill uses a standardized framework for Discoveries, Patterns, and Agent Creation to ensure high-quality, actionable data, but the output can be manually refined in your local repository.

Knowledge Extractor

Name: Knowledge Extractor
Author: rysweet

byrysweet

•

Learning & Documentation

Captures and preserves architectural insights, debugging lessons, and reusable patterns from coding sessions into persistent organizational memory.

The Knowledge Extractor skill transforms ephemeral AI conversations into lasting organizational intelligence by systematically identifying root causes, architectural patterns, and automation opportunities. It analyzes debugging sessions and successful implementations to update local documentation like PATTERNS.md and DISCOVERIES.md, while utilizing a Kuzu memory backend to ensure that hard-won insights from one session are available to inform and improve future development tasks. This skill is essential for teams looking to build a self-improving development environment where failures are documented and successes are standardized.

Key Features

01Structured knowledge templates for DISCOVERIES.md and PATTERNS.md

02Intelligent triggers for suggesting new specialized AI agent creation

03Persistent memory storage via Kuzu backend for cross-session learning

0420 GitHub stars

05Identification and documentation of reusable architectural patterns

06Automated extraction of root causes and solutions from debugging sessions

Use Cases

01Identifying repetitive manual workflows that should be automated with custom agents

02Converting successful one-off implementations into reusable design patterns

03Post-mortem analysis after resolving complex production bugs or CI failures

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rysweet/amplihack knowledge-extractor

For use in Claude.ai and ChatGPT

Download Skill