Education & Research

Lurna

by Lurna

Overview

Turn any topic into study-ready materials directly from ChatGPT. Lurna helps you generate interactive flashcard decks you can flip through in chat, plus structured Markdown study notes in a clean reader view. Review key concepts faster, then export decks and notes to app.lurna.co to keep learning with spaced repetition — making serious review a native part of your AI-study workflow. You can use Lurna app to: 1. Exam prep — flashcards from a lecture topic Use case: Turn a subject you’re studying into an interactive flashcard deck you can review in chat, then export to Lurna for spaced repetition. Example prompt: “Create flashcards for my biology midterm on photosynthesis — cover the light reactions, Calvin cycle, and how ATP and NADPH are used. About 15 cards, exam-level.” 2. Deep-dive study notes from a conversation Use case: After discussing a complex topic with AI, turn the explanation into structured, readable study notes with headings, tables, and summaries Example prompt: “Write study notes on the causes and consequences of the 2008 financial crisis. Include key terms, a timeline, and a short summary I can review before class.” 3. Learn from uploaded materials — notes + flashcards Use case: Upload readings, slides, or your own notes and convert them into study materials without reformatting by hand. Example prompt: “Here are my uploaded lecture notes on CRISPR Cas9. First generate study notes, then create a flashcard set covering the most testable concepts.” 4. Quick review before an interview or certification Use case: Rapidly build recall-focused flashcards for technical interviews, certifications, or vocabulary. Example prompt: “Generate flashcards to help me prep for a AWS Solutions Architect interview — focus on S3, IAM, VPC, and common architecture tradeoffs.”

Tools

create_flashcards

ChatGPT

Build an interactive flashcard deck from AI-generated Q&A pairs and open it in Lurna to study. Show an interactive flashcard deck in the MCP App UI. WORKFLOW (required): 1. When the user asks to create or generate flashcards, compose the study set yourself using the generation rules below. 2. Call this tool with user_prompt, title, description, and flashcards filled from your composed JSON. 3. The server does not generate cards and has no access to notes or external APIs — it only validates and displays what you pass. GENERATION RULES: You are a flashcard generation expert. You must return ONLY valid JSON with no additional text, markdown, or formatting. Invent concrete, accurate, pedagogically useful Q&A pairs appropriate for the topic. Stay grounded in well-established facts; avoid hallucinating obscure citations or fake specifics. If the prompt is vague, choose a reasonable scope for introductory/intermediate study material. Title: {title} Content: {content} Card Design Principles (Apply all of the following when generating cards): 1. Understand-first rule Never create cards for content the learner has not yet studied or that the source does not explain. A flashcard consolidates existing understanding; it does not teach new concepts from scratch. 2. Atomicity Each card tests exactly one atomic piece of knowledge. Every card should be answerable in under 6 seconds during review. If a card would regularly take longer, split it. 3. Production over recognition Cards must require the learner to produce an answer from memory, not merely recognize it. Avoid yes/no or true/false questions. Frame cards so the answer must be generated. 4. Depth of processing Wherever possible, frame questions using "why," "how," or "explain" to force elaborative reasoning rather than rote retrieval. 5. Dual coding When source material contains diagrams, graphs, or spatial information, extract all factual content from them and convert it into text-based cards. 6. Interference management After generating all cards, scan the set for pairs that are likely to be confused (similar terms, similar mechanisms, similar values). For each confusable pair, ensure at least one compare/contrast card exists that directly highlights the distinguishing feature. Card Types (Use a mix of the following types — select whichever best fits each piece of knowledge): 1. Definition cards (forward + reverse): For key terms. Always create both directions (term → definition AND definition → term). 2. Explain/justify cards: For concepts where understanding the reasoning matters. Answers must include the full cause-and-effect chain, not just restate the fact. 3. Compare/contrast cards: For related concepts that are commonly confused. Both sides of the comparison must appear in the answer. 4. Formula/equation cards: For mathematical relationships. Include at least one application card alongside the formula card itself. 5. Enumeration cards: When a set of items exists, prefer individual cards per item over a single list card. Never ask the learner to recall more than 3 items in a single answer. Question framing (use a mix): - "How does X relate to Y?" - "What would happen if...?" - "Compare X and Y" - "What are the implications of...?" - "How would you apply...?" - "Why is X important?" - "What causes / results from...?" - "Explain why [phenomenon occurs]" - "What is the equation for...?" - "[Definition or context] — what term describes this?" Card Rules 1. Generate enough cards to cover the topic well — typically 5–30 cards depending on complexity and source length. 2. Every card must test a distinct concept — no two cards should overlap in what they assess. 3. Answers must be 1–2 sentences maximum, crisp and self-contained. 4. Prioritize high-yield concepts: definitions of foundational terms, relationships, cause-and-effect chains, comparisons, and applications. 5. Do not repeat the question wording inside the answer. 6. Do not create diagram-dependent questions, multi-step numerical calculations, yes/no questions, or cards that list more than 3 items in the a…

generate_notes

ChatGPT

Create formatted Markdown study notes with an A4 reader UI and export them to Lurna. Generate structured Markdown study notes and show them in the MCP App UI (A4 portrait reader). WORKFLOW (required): 1. When the user asks for notes, an outline, or a study guide on a topic, compose the Markdown yourself using the rules below. 2. Call this tool with user_prompt, title, summary, and markdown filled from your composed document. 3. The server does not call an LLM — it validates, renders in the UI, and returns what you pass. GENERATION RULES: You are an expert study-notes writer. Produce clear, accurate, well-structured Markdown suitable for a student learning the topic for the first time or reviewing for an exam. USER REQUEST (verbatim): {user_prompt} SOURCE MATERIAL (if any — notes, paste, outline; otherwise invent from the topic): {content} EXTRA CONSTRAINTS: {constraints} --- Writing principles 1. Accuracy — Use well-established facts. Do not invent citations, page numbers, or paper titles. If unsure, say "commonly described as…" rather than fabricating sources. 2. Clarity — One main idea per paragraph. Define jargon before using it. Prefer plain language; use technical terms only when they add precision. 3. Structure — Use a logical hierarchy: overview → core concepts → mechanisms/details → applications → summary. Headings must reflect how a student would outline the topic. 4. Depth — Match depth to scope: a narrow question gets a focused note; a broad topic gets broader coverage without filler. 5. Active learning — Include brief "Check yourself" prompts or key questions where helpful (not full flashcards). Required document structure Use this outline (adapt section titles to the topic; omit sections that do not apply): ``

markdown [Topic title — clear and specific] > One-sentence hook: why this topic matters. Overview 2–4 sentences: scope, prerequisites, and what the reader will understand after reading. Key terms | Term | Definition | |------|------------| | ... | ... | (Include 5–12 terms for substantial topics; fewer for narrow questions.) Core concepts [Concept 1] Explanation with cause-and-effect where relevant. Use **bold** for critical terms on first use. [Concept 2] ... (Add subsections as needed. Use bullet lists for steps, properties, or comparisons.) How it works / Mechanism Numbered steps or a concise narrative for processes, algorithms, or causal chains. Comparisons & common confusions Use a table or bullet pairs when related ideas are easily mixed up. Applications & examples 1–3 concrete examples tying concepts to real use (science, medicine, engineering, daily life, etc.). Summary - 3–7 bullet takeaways a student should remember. Further study (optional) Topics to explore next — no fake links; plain topic names only.

` Markdown rules - Start with a single # title matching the title field you pass to the tool. - Use ## and ###

 only (no deeper nesting unless essential). - Use tables, bullet lists, and numbered lists where they improve scanability. - Use

inline code

 for symbols, variables, and short formulas; fenced code blocks only for multi-line equations or pseudocode. - Do **not** wrap the entire document in an outer code fence when calling the tool — pass raw markdown in the

markdown

 argument. - Length: roughly 400–2,500 words depending on topic complexity (shorter for narrow questions, longer for broad subjects). Never pad with repetition. Tool call fields When calling

generate_notes, provide: - user_prompt — the user's original request, verbatim - title — matches the H1 heading - summary — 1–2 sentences for previews (not duplicated as the only content) - markdown — the full note body (including # title) - constraints — optional; use "None"` if not specified

Example Prompts

Click any prompt to copy it.

Links

Documentation ↗Privacy Policy ↗

App Stats

Tools

Prompts

ChatGPT

Platforms