Krog - The Complete User Manual & Authoring Guide

1. Introduction & Manifesto

About the Author (Dany)

Hi, I'm Dany. I'm a software engineer currently studying Medicine at the University of El Salvador. Like many students who discovered modern generative AI, I initially threw it at every problem: summarizing clinical textbooks, formatting outlines, and churning out multiple-choice flashcards. But very quickly, a troubling pattern emerged. The tools were undeniably capable, but they were doing the actual thinking for me. Instead of scaffolding my intellectual synthesis, they turned studying into a passive, click-and-skim habit.

Learning requires cognitive friction. In medical physiology, pathology, or systems software engineering, true comprehension is forged when you wrestle with ambiguity, test hypotheses, and confront your own gaps. When software presents a pre-cooked, uneditable summary, it robs you of that synthesis.

Consider NotebookLM: while an impressive technological feat, it operates as a paternalistic black box. You can ask for adjustments, but the underlying system prompt is concealed, the processing logic is proprietary, and the generated documents are treated as immutable artifacts. You cannot click directly into an AI briefing doc to strike through an ambiguous sentence, correct a subtle diagnostic inaccuracy, or reformat an anatomical distinction in your own words.

I built Krog to break that paradigm. Krog is an open, sovereign Learning Management System designed to return complete control of the study process back to the learner. Every prompt is transparent; every generated question is editable; every asset can be stored, exported, and run on the user's terms—whether through cloud APIs or local models running offline on your own machine.

What Krog Is (and What It Is Not)

• Krog IS: An environment for active recall, modular content authoring, structured evaluation, real-time peer learning, and rigorous institutional assessments.
• Krog IS NOT: An opaque wrapper around a closed API that locks your knowledge behind paywalls and proprietary formats.

Core Architectural Primitives

The Krog ecosystem is built upon six interconnected primitives:

1. Contents (Content): Long-form markdown notes, uploaded PDF references, document extracts, and web resources.
2. Drawings (Drawing): Vector whiteboard diagrams, visual mind maps, and anatomical sketches.
3. Questions (Question): Atomic active-recall items adhering to a strict, portable schema.
4. Exams (Exam): Cohesive suites of questions paired with optional study material, time parameters, and access controls.
5. Collections (Collection): Curated paths grouping multiple exams and study materials into semester-long curricula.
6. Organizations (Organization): Multi-tenant institutional hubs featuring teams, shared vaults, and granular permissions.

2. Account, Personalization & Visual Flavors

Profile & Identity

Every user on Krog has a unique handle (username), full name, bio, and customizable avatar.

• Public Visibility Toggle: You can independently choose whether your profile, social activity, and study statistics are public or restricted strictly to your organization and followers (profile_picture_public, social_public).
• Study Streaks & Milestones: Daily consistency is tracked via streak and longestStreak. Completing a daily retrieval drill or reading assignment maintains the streak counter.
• Badges: Performance achievements unlock profile badges.

The Flavor System: 50+ Curated Typographic Themes

Visual fatigue during long study sessions contributes heavily to cognitive burnout. Krog features an extensive theme engine known as Design Flavors. Rather than generic dark/light modes, Flavors alter palette variables, contrast ratios, border radiuses, and typographic rendering:

• Minimalist & Paper: minimal, notebook, legalpad, chalkboard, typewriter, eink, newspaper.
• Technical & Dark: krog (signature emerald-dark), nord, dracula, cyberpunk, terminal, developer, repository, charcoal, onyx.
• Artistic & Warm: swiss, bauhaus, midcentury, sakura, warmcream, coffee, desert, sumi, kyoto.
• Focus & High Contrast: studyfocus, highcontrast, scientific, blueprint, wireframe.

Users can also independently customize the App Font (UI controls) and the Material Font (long-form reading serif, such as Literata or Merriweather).

3. The Authoring Engine: Writing Study Material in Krog

Writing study material in Krog is built on the concept of the Executable Textbook. Rather than producing static, passive text, Krog's Markdown engine empowers authors to construct dynamic, interactive pedagogical documents using plain text directives, structured semantic containers, and embedded active-recall widgets.

Every section below showcases its feature using interactive "Text" and "Result" tabs so you can inspect the exact authoring markup and immediately see how Krog renders it.

1. Chapter Boundaries (------)

Study material can be split into navigable chapters by inserting six dashes on their own line:

Copy code
# Chapter 1: Foundations of Neurotransmission
Content for chapter 1...

------

# Chapter 2: Synaptic Plasticity & Long-Term Potentiation
Content for chapter 2...

Krog parses each chapter into an interactive reading view with a sidebar table of contents, reading time estimates, and per-chapter completion checkboxes (StudyProgress). Flow directives declared in preceding chapters automatically carry forward across boundaries.

2. Flow Directives Reference

Flow directives begin with @ at the start of a line. They modify document layout and typography dynamically from that point downward until changed again or reset:

To reset any directive to its default, pass default, off, or none (e.g. @align default, @columns off).

3. Inline Enhancements: Colors, Highlights & Typography

Copy code
* **Scoped Fonts**: <font jbmono>Monospaced code snippet</font> and <font cmu>Mathematical CMU Serif</font>.
* **Size Spans**: <size 20>Prominent note text</size>.
* **Color Tags**: <c emerald>Emerald highlight</c> and <c rose>Rose critical text</c>.
* **Fluorescent Highlights**: ==High-yield concept== and =={sky}:Sky-tinted phrase==.
* **Arrow Auto-Formatting**: Resting state → Depolarization ⇒ Action potential threshold.

4. Active Recall: Cloze Deletions, Spoilers & Speech

Click-to-Reveal Cloze Deletions

Wrap words in double curly braces to hide them behind an interactive clickable pill:

Copy code
The action potential in ventricular myocytes begins with Phase 0, driven by the rapid opening of {{voltage-gated fast Na+ channels}}.

Interactive Spoiler Blocks (<spoiler>)

Hide full diagnostic solutions or rationales until clicked:

Copy code
What is the antidote for acetaminophen toxicity?
<spoiler>
**N-acetylcysteine (NAC)**, which replenishes hepatic glutathione stores.
</spoiler>

Text-to-Speech Pronunciation (<play>)

Clicking the word speaks it out loud using the browser's speech synthesis engine:

Copy code
Verify medical terminology: <play lang="es">esternocleidomastoideo</play> and French anatomy: <play lang="fr">trompe-l'œil</play>.

5. Semantic Container Callouts

Wrap contextual notes inside semantic XML tags. Unclosed tags safely auto-close at the next heading:

Copy code
<tip>
**Mnemonic for Essential Amino Acids:** PVT TIM HaLL
(Phenylalanine, Valine, Threonine, Tryptophan, Isoleucine, Methionine, Histidine, Leucine, Lysine).
</tip>

<warning>
Beta-blockers without intrinsic sympathomimetic activity are strictly contraindicated in acute decompensated heart failure with pulmonary edema.
</warning>

<key>
The primary hemodynamic hallmark of severe aortic stenosis is a systolic ejection pressure gradient between the left ventricle and the aorta.
</key>

6. Academic & Scientific Numbered Boxes

For formal STEM curricula, Krog features numbered theorem environments with automatic CSS counter numbering:

Copy code
<theorem Central Limit Theorem>
Let $X_1, X_2, \dots, X_n$ be independent and identically distributed random variables with $\mathbb{E}[X_i] = \mu$ and $\operatorname{Var}(X_i) = \sigma^2 < \infty$. 
As $n \to \infty$, the sample mean $\bar{X}_n$ converges in distribution to a normal distribution:
$$\sqrt{n}\left(\bar{X}_n - \mu\right) \xrightarrow{d} \mathcal{N}(0, \sigma^2)$$
</theorem>

<example Clinical Vignette>
A 62-year-old male with sudden onset retrosternal crushing chest pain radiating to the left jaw. Troponin I is elevated at 4.2 ng/mL. The most urgent next step is activating the cardiac catheterization team for primary percutaneous coronary intervention (PCI).
</example>

7. Step-by-Step Clinical & Procedural Workflows (<steps>)

Format sequential clinical workflows or engineering algorithms with high-visibility numbered step badges:

Copy code
<steps>
1. **Initial Assessment**: Verify airway patency, breathing stability, and circulation (ABCs).
2. **Access & Labs**: Establish two large-bore peripheral IVs; draw type and screen.
3. **Fluid Resuscitation**: Administer weight-based crystalloid bolus (30 mL/kg).
4. **Re-evaluation**: Assess mean arterial pressure (MAP) and hourly urine output response.
</steps>

8. Architectural & Clinical Process Diagrams (<mermaid>)

Krog natively compiles Mermaid diagrams via fenced code blocks or the <mermaid> tag: