Where are these learned skills stored?

Skills are stored locally in your project repository under the .claude/skills/[skill-name]/ directory as SKILL.md files.

How does Claude use these skills later?

Because the skills are stored in a standardized format with specific trigger conditions, Claude can read the skills directory in future sessions to automatically apply previously learned solutions.

When is the best time to invoke this skill?

Invoke it when you encounter non-obvious debugging fixes, misleading error messages, or multi-step processes that aren't well-documented elsewhere.

What does the /learn command actually do?

It triggers a workflow where Claude analyzes recent discoveries, checks for existing documentation, and creates a new .claude/skills/ file containing the problem, context, and solution.

Knowledge Extraction & Skill Builder

Name: Knowledge Extraction & Skill Builder
Author: kirbyh

bykirbyh

0•

Productividad y Flujo de Trabajo

Extracts non-obvious insights and complex workflows into persistent, reusable skills for future Claude Code sessions.

The Learn skill provides a structured framework for Claude to formalize its own problem-solving experiences, transforming temporary session insights into long-term project assets. By identifying non-obvious debugging solutions, complex tool integrations, or creative workarounds, this skill ensures that hard-won knowledge is documented and accessible in future interactions. It guides the AI through a rigorous evaluation phase to verify the value of the discovery, checks for existing skills to prevent redundancy, and generates standardized documentation that allows Claude to recognize specific triggers and apply solutions instantly when similar issues arise.

Características Principales

01Automated knowledge extraction from active sessions

02Duplicate detection and versioning for existing skills

030 GitHub stars

04Specific trigger-condition mapping for future recall

05Standardized markdown templates for skill persistence

06Self-assessment logic to filter for high-value insights

Casos de Uso

01Capturing multi-step data processing or deployment workflows

02Documenting undocumented library workarounds or API quirks

03Formalizing complex debugging paths for misleading error messages

Características Principales

01Automated knowledge extraction from active sessions

02Duplicate detection and versioning for existing skills

030 GitHub stars

04Specific trigger-condition mapping for future recall

05Standardized markdown templates for skill persistence

06Self-assessment logic to filter for high-value insights

Casos de Uso

01Capturing multi-step data processing or deployment workflows

02Documenting undocumented library workarounds or API quirks

03Formalizing complex debugging paths for misleading error messages