rb:learn

Capture and store lessons from bug fixes with duplicate detection and consent.

7|Updated Mar 22, 2026
One-click install
npx skills add https://github.com/slbug/claude-ruby-grape-rails --skill rb-learn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rb:learn
Source: https://github.com/slbug/claude-ruby-grape-rails/tree/main/plugins/ruby-grape-rails/skills/learn
Command: npx skills add https://github.com/slbug/claude-ruby-grape-rails --skill rb-learn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps capture and preserve actionable lessons when a bug is fixed or a correction is provided, preventing repeated mistakes across the codebase and future projects.

Core Features & Use Cases

  • Root cause extraction: Classifies fixes as knowledge gaps, wrong assumptions, tooling gaps, or process failures to capture the real root cause rather than symptoms.
  • Duplicate detection: Searches shipped references, project CLAUDE.md, .claude/solutions, and auto-memory to avoid redundant entries.
  • Destination guidance and consent: Recommends and writes concise rules to project CLAUDE.md, prompts for consent before adding cross-project auto-memory, or hands complex stories to .claude/solutions.

Quick Start

Run /rb:learn with a short description of what was fixed and the root cause to record the lesson and choose the appropriate destination.

Frequently Asked Questions about rb:learn

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I capture lessons from Rails bug fixes to prevent repeating mistakes?

Capture lessons from Rails bug fixes by providing the fix description and root cause to classify the mistake and store it as a reusable rule in project CLAUDE.md or cross-project auto-memory.

What is the best way to document ActiveRecord corrections across multiple projects?

Document ActiveRecord corrections by extracting the root cause, checking for duplicate entries, and prompting for user consent before writing the lesson to cross-project auto-memory.

How does duplicate detection work when saving Sidekiq troubleshooting lessons?

Duplicate detection for Sidekiq lessons works by searching existing shipped references, project CLAUDE.md, .claude/solutions, and auto-memory to avoid creating redundant entries before saving a new lesson.

Can I use this approach to categorize Grape API process failures and tooling gaps?

Yes, categorize Grape API process failures and tooling gaps by classifying fixes into knowledge gaps, wrong assumptions, tooling gaps, or process failures to capture the true root cause rather than symptoms.

When should I hand off complex bug fixes to .claude/solutions instead of CLAUDE.md?

Hand off complex bug fix stories to .claude/solutions instead of project CLAUDE.md when the lesson is too detailed for a concise rule, ensuring CLAUDE.md remains focused on direct project guidelines.