learn

Persist new knowledge into agent system files and skill definitions.

1|Updated Mar 20, 2026

Search Tags:#skills #agents #docs #system #learn #knowledge-persistence

Ready to run

npx skills add https://github.com/rupertsworld/skills --skill learn-rupertsworld

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: learn
Source: https://github.com/rupertsworld/skills/tree/main/learn
Command: npx skills add https://github.com/rupertsworld/skills --skill learn-rupertsworld

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Persist new knowledge into the agent system. Covers everything from "remember my email" to "create a new skill." Refer to any provided documentation for system architecture and conventions before making structural changes.

Core Features & Use Cases

Persist facts: update AGENTS.md, docs, and the skill's SKILL.md
Create or modify skills under skills/<name>/SKILL.md
Reference architecture docs before applying changes

Quick Start

Persist a new fact by updating AGENTS.md or docs, then optionally create or modify skills under skills/<name>/SKILL.md.

Frequently Asked Questions about learn

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

FAQPage Schema

How do I persist new knowledge and update agent system facts?▼

To persist new knowledge, you update identity facts, preferences, or system rules by modifying AGENTS.md and docs. This ensures your agent system retains the latest information and operates with the correct configuration.

Can I create or modify skills by updating the SKILL.md file?▼

Yes, you can create or modify skills by writing to the skills/<name>/SKILL.md file path. This allows you to directly add new capabilities or adjust existing agent behaviors within the system architecture.

How does knowledge persistence work for updating agent preferences and system rules?▼

Knowledge persistence works by reading target files before applying non-destructive updates. It ensures that changes to agent preferences, identity facts, and system rules are safely integrated without overwriting existing documented architecture.

Do I need to reference architecture docs before making structural changes to the system?▼

Yes, you must reference any provided architecture documentation before making structural changes. This ensures modifications to skills and system files adhere to established conventions and prevents architectural inconsistencies.

What are the limitations when applying non-destructive updates to agent skills?▼

The primary limitation is the requirement to read target files before applying non-destructive updates. You cannot blindly overwrite files; modifications must respect existing documented architecture guidelines and system conventions.