Claude Code Plugins

Community-maintained marketplace

Feedback

A learning methodology skill based on Dr. Justin Sung's PACER system for effective information retention and application. Use when helping users (1) learn new material efficiently, (2) take notes or summarize content, (3) create study plans, (4) review or memorize information, (5) understand how to retain knowledge better, or (6) process educational content like books, lectures, courses, or articles.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name pacer-learning-system
description A learning methodology skill based on Dr. Justin Sung's PACER system for effective information retention and application. Use when helping users (1) learn new material efficiently, (2) take notes or summarize content, (3) create study plans, (4) review or memorize information, (5) understand how to retain knowledge better, or (6) process educational content like books, lectures, courses, or articles.

PACER Learning System

A systematic approach to learning based on cognitive science principles, developed by Dr. Justin Sung.


Part 1: First Principles — How Learning Actually Works

1.1 The Two-Stage Model of Learning

Learning consists of two distinct stages:

Stage Definition Common Mistake
Consumption Intake of information (reading, listening, watching) People focus almost entirely here
Digestion Processing, understanding, and encoding into long-term memory Almost completely neglected

First Principle: Consumption without digestion = forgetting. Studies suggest up to 90% of consumed information is forgotten without proper digestion.

1.2 Memory Architecture: STM → Encoding → LTM

How information flows in the brain:

Input → Short-Term Memory (STM) → [Encoding] → Long-Term Memory (LTM)
              ↓                                         ↓
        Forgotten in seconds                    Retained & retrievable
        (capacity: ~7 items)                    (effectively unlimited)

First Principle: Information in STM is lost within seconds unless actively encoded into LTM. The encoding process—not mere exposure—determines retention.

1.3 Encoding Quality Depends on Connections

From Dr. Justin Sung: "The brain's ability to construct memory is called encoding and it depends a lot on the connections you can make with new information. When you build more connections, you can retain the information better."

First Principle: Isolated facts are hard to retain. Information connected to existing knowledge networks encodes more strongly and retrieves more easily.

1.4 The Kim Peek Paradox: Memory ≠ Understanding

Kim Peek (the "Rain Man" savant) could memorize entire books verbatim but struggled with reasoning and problem-solving.

First Principle: Perfect recall is neither possible nor desirable for most learners. The goal is to retain what's needed in a form that enables application, not just reproduction.

1.5 The Overeating Analogy

Consuming information without digestion is like eating without digestion—you will "vomit" (forget) everything.

First Principle: The two stages MUST be balanced. Time spent consuming must be matched with time spent digesting. When time is limited, reduce consumption rather than skip digestion.

1.6 Expert Knowledge is Networked, Not Linear

Textbooks present information linearly (sentence after sentence), but expert knowledge exists as highly interconnected networks. Experts can start at any node and navigate to any other.

First Principle: The learner's job is to reconstruct the author's mental network, not to memorize the linear sequence of words.

1.7 Different Information Types Require Different Processes

Using the same process (e.g., re-reading, highlighting) for all information types is ineffective.

First Principle: Categorize information by type, then apply the specific encoding strategy optimized for that type.


Part 2: The PACER Framework

2.1 Overview

Categorize information by type, then apply the matching encoding process:

Type Definition Process Key Action
Procedural How to do something (steps, techniques, syntax) Practice Apply immediately in real context
Analogous Relates to existing knowledge Critique Examine: How similar? How different? When does analogy break?
Conceptual What something is (facts, theories, principles, relationships) Map Create non-linear mind maps showing connections
Evidence Details that make concepts concrete (dates, statistics, cases) Store & Rehearse Record → later apply via problems/teaching
Reference Nitty-gritty details needed for recall (constants, names, lists) Store & Rehearse Record → use spaced repetition (e.g., Anki)

2.2 Detailed Processing Guidelines

P - Procedural Information

Identify by: Instructions, steps, syntax, "how to" content.

Process - Practice:

  • Apply as soon as possible after consumption
  • Do NOT memorize first then practice later—you'll forget
  • If no time to practice now, stop consuming; don't waste time on rote memorization

Examples: Coding syntax, clinical examination techniques, language grammar, cooking methods

A - Analogous Information

Identify by: Content that reminds you of something you already know.

Process - Critique:

  1. Explicitly identify the connection to prior knowledge
  2. Ask: In what specific ways are these similar?
  3. Ask: In what ways are they different?
  4. Ask: When/where does this analogy break down?
  5. Ask: Is there a better or more comprehensive analogy?

Why it works: Extends your existing knowledge network rather than creating isolated new nodes.

C - Conceptual Information

Identify by: Definitions, explanations, theories, principles, "what" content.

Process - Map:

  • Create non-linear, network-based notes (mind maps)
  • Focus on connections between concepts, not linear sequences
  • Expert knowledge exists as networks, not sentences—recreate that structure
  • Add to and reorganize your map as you read more

Key insight: Textbooks present linearly, but expert knowledge is networked. Your job is to reconstruct the author's mental network.

E - Evidence Information

Identify by: Specific examples, statistics, case studies, historical events with dates/names/places.

Process - Store & Rehearse:

  • Store: Immediately record in a second brain (Notion, Obsidian, flashcards, or your concept map)
  • Rehearse (later): Apply through problem-solving, essay writing, teaching, or explaining
  • Do NOT stop to memorize while reading—that steals time from P, A, C processing

R - Reference Information

Identify by: Highly specific details not crucial for understanding (constants, gene names, exact values).

Process - Store & Rehearse:

  • Store: Dump into flashcard system immediately
  • Rehearse: Use spaced repetition (e.g., Anki) during dedicated review time
  • Absolutely do NOT waste consumption time trying to memorize these on the spot

Part 3: Critical Rules (Derived from First Principles)

  1. Balance is mandatory: Every unit of consumption requires corresponding digestion. Overconsuming = mental vomiting (forgetting).

  2. Don't fight biology: The brain has limits on how much can be encoded at once. Respect them.

  3. P, A, C are foundational: These three build the knowledge structure. E and R attach to that structure. Skipping P/A/C to memorize E/R is backwards.

  4. Time pressure trap: When short on time, people consume more. This is wrong. Consume less, digest properly.

  5. Passive reading detector: If you reach the end of a page and can't recall what you read, you've entered passive mode. Stop and digest.


Part 4: Application Workflow

When helping users learn:

  1. Identify the information type — Ask: Is this P, A, C, E, or R?
  2. Apply the matching process — Guide them to practice, critique, map, or store+rehearse
  3. Enforce balance — If they're consuming too fast, slow them down
  4. Prioritize P, A, C — These form the foundation; E and R build on top

Part 5: Example Application

User wants to learn about machine learning:

Content Type Action
"How to implement gradient descent" P (Procedural) Write code immediately; don't just read
"This is like how water flows downhill" A (Analogous) Critique: How is optimization like gravity? Where does analogy fail?
"Neural networks consist of layers..." C (Conceptual) Map the relationships: layers → neurons → weights → activations
"ImageNet has 14M images, 20K categories" E (Evidence) Store; later use when explaining dataset scales
"Learning rate typically 0.001-0.01" R (Reference) Flashcard; review via spaced repetition

Part 6: When Creating Study Materials

Structure content to support proper digestion:

  • For tutorials: Include practice exercises immediately after each concept
  • For summaries: Organize as concept maps, not linear bullet points
  • For flashcards: Reserve for E and R types only; never for P, A, or C
  • For study plans: Allocate 50%+ time to digestion activities, not just reading