improve-skill

Refine or create agent skills from observed conversation evidence.

35|Updated Sep 20, 2022
One-click install
npx skills add https://github.com/Andriiklymiuk/corgi --skill improve-skill-andriiklymiuk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: improve-skill
Source: https://github.com/Andriiklymiuk/corgi/tree/main/plugins/corgi/skills/improve-skill
Command: npx skills add https://github.com/Andriiklymiuk/corgi --skill improve-skill-andriiklymiuk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill allows users to improve or create new agent skills based on observed gaps or friction points within a conversation, promoting continuous skill refinement and adaptation.

Core Features & Use Cases

  • Skill Improvement: Refine existing skills by addressing specific gaps identified during conversations.
  • Skill Creation: Scaffold new skills from the evidence of a conversation or a workflow, enabling quick adaptation to new scenarios.
  • Use Case: If a conversation reveals a gap in an existing skill, the user can use this Skill to refine the skill or create a new one based on the conversation evidence.

Quick Start

Run the 'improve-skill' command followed by the evidence or the SKILL.md file to refine or create a new skill.

Frequently Asked Questions about improve-skill

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

FAQPage Schema
How do I refine an existing AI skill using conversation evidence?

You can refine an existing AI skill by running the 'improve-skill' command and providing the observed conversation evidence or the SKILL.md file to address specific gaps identified during the interaction.

Can I create a new agent skill from a workflow or chat history?

Yes, you can scaffold new agent skills from conversation evidence. The skill definition requires real observed gaps and user intent to ensure the new skill adapts to the scenario.

What is the best way to adapt agent skills for new scenarios?

The best way to adapt agent skills is to use conversation evidence to identify friction points, then apply evidence-driven improvements to refine the skill or create a new one.

Does skill refinement require a specific environment or dependencies?

No specific dependencies are required. You can execute the 'improve-skill' command directly with your conversation evidence or SKILL.md file to generate the refined output.

Why should I avoid one-off stack quirks when creating agent skills?

Avoiding one-off stack quirks ensures your agent skills remain robust and generally applicable. The skill creation process relies on real observed gaps and user intent rather than isolated anomalies.