teach

name	teach
description	Explain complex concepts with analogies, visual metaphors, recursive diagnostics, and PhD-level depth using consulting frameworks
triggers	["teach","teach me","deep explain","break down concept"]
argument-hint	<topic or concept> [--ko for Korean]

Teach Skill

Purpose

Explain complex concepts in a detailed report by integrating intuitive analogies, visual metaphors, and recursive diagnostics to ensure clarity and understanding.

When to Activate

• User asks to understand, learn, or be taught a concept
• User says "teach me", "explain", "break down", "help me understand"
• User asks "how does X work?" or "what is X?" in a learning context
• Any request where deep conceptual explanation is the goal

Workflow

1. Identify Key Concepts

Outline the main ideas or concepts within the topic. Start with the big picture or a high-level analogy, then gradually zoom in to subcomponents.

• Map the concept space using MECE (Mutually Exclusive, Collectively Exhaustive) decomposition
• Identify prerequisite knowledge the reader needs
• Establish a logical progression from simple to complex

2. Explain with Analogies

For each key concept, break it down using intuitive analogies to create relatable visual anchors.

• Choose analogies from everyday experience (kitchen, traffic, building construction, etc.)
• Ensure analogies are structurally accurate, not just superficially similar
• Layer analogies: start simple, then extend the metaphor as complexity increases

3. Research Supporting Details

Gather quantitative data, statistics, and facts. Relate these back to the analogies.

• Use concrete numbers and real-world benchmarks
• Reference seminal papers, key researchers, or foundational texts
• Connect data points to the analogies established in Step 2
• When WebSearch or exa-search tools are available, use them to verify data and find current statistics; otherwise rely on training knowledge and flag uncertain claims with "(approximate)" or "(as of [year])"

4. Recursive Understanding Check (Embedded)

Embed brief comprehension probes directly into the written report after each major section — these are part of the document, not interactive pauses.

• After each key concept section, include a "Check Your Understanding" subsection with 1-2 targeted questions
• If a prerequisite concept is complex, expand on it inline before continuing
• Use Socratic-style questions that reveal whether the reader grasped the deep structure, not just surface facts

5. Draft the Report

Integrate each key concept with its analogies and supporting details for clarity.

• Use Markdown structure: # H1 for the topic, ## H2 for major sections, ### H3 for subsections
• Include diagrams, tables, or visual metaphors where they aid understanding
• Maintain PhD-level depth while keeping language accessible

6. Zoom Out and Reconnect

Re-explain the original concept from the top, ensuring mastery of all subcomponents.

• Show how all the pieces fit together
• Revisit the opening analogy with enriched understanding
• Highlight connections between subcomponents that weren't obvious at first

7. Review for Clarity and Accuracy

Ensure all details are accurate, properly referenced, and clearly presented.

• Verify factual claims and statistics
• Check that analogies don't mislead or oversimplify critically
• Ensure logical flow from section to section

Format Rules

• Always use advanced consulting techniques (e.g., MECE, issue trees, hypothesis-driven structure)
• Derive long, detailed, PhD-level in-depth explanations and insights
• Format in Markdown: # H1, ## H2, ### H3, etc.
• Default language: English. Pass --ko to write in Korean instead
• Cite sources in APA format (Author, Year) with full references at the end — only cite well-known foundational works you are confident exist; flag any uncertain citations with "[verify]"
• End with a technical comprehension question to determine the reader's mastery and suggest next steps

Examples

/teach transformer architecture
/teach how does backpropagation work
/teach explain RLHF in large language models
/teach what is knowledge distillation

Notes

• Prioritize structural accuracy of analogies over simplicity — a slightly complex but correct analogy beats a simple but misleading one
• When the topic spans multiple disciplines, use MECE to ensure no blind spots
• For highly technical topics, don't shy away from math — but always pair equations with intuitive explanations
• The final comprehension question should be diagnostic: it should reveal whether the reader understood the deep structure, not just surface facts