anki-card-creator

name anki-card-creator description Convert medical textbook content, lecture notes, and study materials into Anki flashcards using spaced repetition optimization. Supports multiple card types (basic, cloze, image occlusion) with automated tagging and deck organization for efficient medical exam preparation. allowed-tools ["Read","Write","Bash","Edit"] license MIT metadata {"skill-author":"AIPOCH"} Anki Card Creator Overview Intelligent flashcard generation tool that transforms medical study materials into Anki-compatible cards optimized for long-term retention through evidence-based spaced repetition. Key Capabilities: Multi-Format Input : PDF textbooks, lecture slides, notes, web articles Card Type Selection : Basic, reversible, cloze deletion, image occlusion Spaced Repetition Optimization : Cards formatted for optimal review intervals Automated Tagging : Hierarchical organization by subject and topic Media Integration : Auto-download and embed relevant images Batch Processing : Convert entire chapters or courses efficiently When to Use ✅ Use this skill when: Preparing for medical board exams (USMLE, MCAT, specialty boards) Memorizing drug names, mechanisms, and side effects Learning anatomy structures and relationships Retaining biochemical pathways and reactions Creating pathology differential diagnosis cards Converting lecture notes to active recall format Building comprehensive review decks from textbooks ❌ Do NOT use when: Learning conceptual understanding → Use textbooks/videos first Memorizing without context → Ensure comprehension before card creation Creating cards for rapidly changing information → Use current resources Substituting clinical experience → Cards supplement, not replace practice Complex procedural skills → Use simulation or hands-on training Integration: Upstream : abstract-summarizer (textbook chapter condensation), pdf-text-extractor (content extraction) Downstream : anki-sync-server (cloud backup), study-schedule-optimizer (review planning) Core Capabilities 1. Cloze Deletion Generation Create fill-in-the-blank cards from dense text: from scripts.card_creator import AnkiCardCreator creator = AnkiCardCreator() # Generate cloze cards from text text = """ Acute inflammation is characterized by five cardinal signs: redness (rubor), heat (calor), swelling (tumor), pain (dolor), and loss of function (functio laesa). These result from vasodilation, increased vascular permeability, and leukocyte infiltration. """ cards = creator.create_cloze_cards( text=text, n_cards= 3 , # Generate 3 cards from this text difficulty= "intermediate" , tags=[ "Pathology" , "Inflammation" , "General_Pathology" ] ) # Output format (Anki-compatible) # Card 1: "Acute inflammation is characterized by {{c1::five}} cardinal signs..." # Card 2: "...signs: {{c1::redness}} (rubor), {{c2::heat}} (calor)..." Cloze Strategies: Strategy Use Case Example Single Deletion Key facts "The {{c1::liver}} is the largest internal organ" Multiple Deletions Lists/groups "CRAB symptoms: {{c1::Calcium}}, {{c2::Renal}}, {{c3::Anemia}}, {{c4::Bone}}" Hierarchical Progressive detail "{{c1::Metformin}} is a {{c2::biguanide}} that {{c3::decreases hepatic glucose production}}" 2. Basic and Reversible Cards Generate question-answer pairs: # Create basic cards cards = creator.create_basic_cards( content={ "What is the mechanism of action of penicillin?" : "Inhibition of bacterial cell wall synthesis by binding to PBPs" , "What are the major side effects of ACE inhibitors?" : "Dry cough, hyperkalemia, angioedema, hypotension" }, reversible= True , # Create reverse cards too tags=[ "Pharmacology" , "Antibiotics" ] ) Card Types: Basic : Question → Answer Reversible : Both directions (Q→A and A→Q) Type-in : Student types answer for active recall Hint Cards : Progressive disclosure with hints 3. Image Occlusion Cards Create visual learning cards from diagrams: # Image occlusion for anatomy cards = creator.create_image_occlusion( image_path= "brachial_plexus.png" , occlude_regions=[ { "label" : "Musculocutaneous nerve" , "coords" : ( 120 , 200 , 200 , 240 )}, { "label" : "Median nerve" , "coords" : ( 300 , 250 , 380 , 290 )}, { "label" : "Ulnar nerve" , "coords" : ( 400 , 280 , 480 , 320 )} ], card_type= "one_by_one" # or "all_at_once" ) Occlusion Types: One by One : Reveal structures progressively All at Once : Identify all occluded regions Transparent : Hint visible through occlusion Colored : Different colors for different structures 4. Smart Tagging and Organization Automatically organize cards into logical hierarchies: # Auto-tag based on content cards = creator.process_textbook_chapter( chapter_text=pathoma_chapter, source= "Pathoma - Hematopoiesis" , auto_tag= True , tag_hierarchy=[ "Pathoma" , "{{chapter_name}}" , "{{section_name}}" ] ) # Results in tags like: # Pathoma::Hematopoiesis::Red_Cell_Disorders # Pathoma::Hematopoiesis::White_Cell_Disorders Tagging Strategies: Source-based : First Aid, Pathoma, Sketchy Subject-based : Anatomy, Physiology, Pharmacology Exam-based : USMLE_Step1, Step2_CK, Shelf_Exams Difficulty : High_Yield, Medium_Yield, Low_Yield Common Patterns Pattern 1: Textbook Chapter Conversion Scenario : Convert entire First Aid chapter to Anki deck. # Extract and convert chapter python scripts/main.py \ --input first_aid_cardiology.pdf \ --chapter "Ischemic Heart Disease" \ --card-type cloze \ --n-cards-per-page 3 \ --tags "First_Aid::Cardiology::IHD" \ --output ischemic_heart_disease.apkg # Import directly to Anki # File ready for import with media included Workflow: Read chapter for understanding first Run conversion tool Review generated cards for accuracy Import to Anki Study with standard spaced repetition Pattern 2: Drug Cards with Images Scenario : Create comprehensive pharmacology cards. # Generate drug cards with mechanisms drugs = [ { "name" : "Metformin" , "class" : "Biguanide" , "mechanism" : "Activates AMPK, decreases hepatic gluconeogenesis" , "indications" : "Type 2 diabetes, PCOS" , "side_effects" : "GI upset, lactic acidosis (rare), B12 deficiency" , "image" : "metformin_structure.png" } ] cards = creator.create_drug_cards( drugs=drugs, include_structures= True , include_mechanism_diagrams= True , tags=[ "Pharmacology" , "Diabetes" , "First_Line" ] ) Card Format: Front: Metformin - Class? Back: Biguanide Tags: Pharmacology::Diabetes::First_Line Front: Metformin - Mechanism? Back: Activates AMPK → decreases hepatic gluconeogenesis Image: [mechanism diagram] Tags: Pharmacology::Diabetes::Mechanism Front: Metformin - Side effects? Back: GI upset, lactic acidosis (rare), B12 deficiency Tags: Pharmacology::Diabetes::Side_Effects Pattern 3: Pathology Differential Cards Scenario : Create cards for differential diagnosis practice. # Differential diagnosis cards cards = creator.create_differential_cards( condition= "Cough with hemoptysis" , differentials=[ { "diagnosis" : "Lung cancer" , "key_features" : "Smoker, weight loss, mass on CT" }, { "diagnosis" : "TB" , "key_features" : "Immigrant, night sweats, cavitary lesion" }, { "diagnosis" : "PE" , "key_features" : "Sudden onset, tachypnea, D-dimer elevated" }, { "diagnosis" : "Bronchiectasis" , "key_features" : "Chronic cough, purulent sputum, dilated airways" } ], card_type= "compare_contrast" ) Card Types: Feature → Diagnosis : "Smoker + hemoptysis + weight loss → ?" Diagnosis → Features : "Lung cancer presentation?" Compare : "Differentiate lung cancer vs TB" Pattern 4: Lecture Notes to Cards Scenario : Convert professor's lecture slides to study cards. # Process lecture PDF python scripts/main.py \ --input lecture_slides.pdf \ -- source "Dr_Smith_Cardiology_Lecture_5" \ --card-type mixed \ --extract-images \ --auto-tag \ --output cardiology_lecture_5.apkg # Review and edit in Anki # Delete low-yield cards # Add personal mnemonics Post-Processing Tips: Delete obvious or overly simple cards Merge redundant cards Add personal memory hooks Suspend low-yield topics Complete Workflow Example Building comprehensive USMLE Step 1 deck: # Step 1: Process multiple sources python scripts/main.py \ --input first_aid_pathology.pdf \ -- source "First_Aid_2024" \ --card-type cloze \ --tags "First_Aid::Pathology" \ --output fa_pathology.apkg python scripts/main.py \ --input sketchy_pharm.pdf \ -- source "Sketchy_Pharm" \ --card-type basic \ --include-images \ --tags "Sketchy::Pharm" \ --output sketchy_pharm.apkg # Step 2: Merge decks python scripts/main.py \ --merge fa_pathology.apkg sketchy_pharm.apkg \ --output usmle_step1_master.apkg # Step 3: Add leech tags for difficult cards python scripts/main.py \ --input usmle_step1_master.apkg \ --tag-leeches \ --leech-threshold 8 \ --output usmle_step1_tagged.apkg Python API: from scripts.card_creator import AnkiCardCreator from scripts.importers import PDFImporter # Initialize creator = AnkiCardCreator() importer = PDFImporter() # Import and process textbook content = importer.import_pdf( path= "robbins_pathologic_basis_of_disease.pdf" , pages= "150-200" , # Inflammation chapter extract_images= True ) # Create cloze cards cards = creator.create_cloze_cards( text=content.text, images=content.images, n_cards= 50 , difficulty= "advanced" , tags=[ "Robbins" , "General_Pathology" , "Inflammation" ] ) # Add image occlusion for diagrams for diagram in content.diagrams: occlusion_cards = creator.create_image_occlusion( image=diagram, auto_detect_labels= True ) cards.extend(occlusion_cards) # Export Anki package creator.export_apkg( cards=cards, deck_name= "Robbins_Inflammation" , output_path= "robbins_inflammation.apkg" ) print ( f"Created { len (cards)} cards ready for Anki import" ) Quality Checklist Card Quality: One fact per card (atomicity) Clear, unambiguous questions Answers specific and complete Cloze deletions don't give away answer Images high resolution and relevant Tags hierarchical and consistent Educational Value: Cards test high-yield information Difficulty appropriate for level Connections between related concepts Clinical context included where relevant No duplicate or near-duplicate cards Technical Quality: Valid Anki format (import tested) Media files properly linked UTF-8 encoding for special characters CSS styling readable on mobile Card types appropriate for content Before Study: CRITICAL : Review cards for accuracy CRITICAL : Delete cards for mastered material Personalize with own mnemonics Adjust card order if needed Set appropriate daily review limits Common Pitfalls Content Issues: ❌ Too much information → "Cramming" cards with multiple facts ✅ One concept per card; break complex topics into multiple cards ❌ Memorization without understanding → Memorizing without context ✅ Include "why" and clinical significance ❌ Outdated information → Old guidelines or disproven theories ✅ Verify against current textbooks and guidelines Card Design Issues: ❌ Ambiguous clozes → "The {{c1::liver}} produces {{c2::bile}}" (which one is being asked?) ✅ Clear question: "The liver produces ?" or "What organ produces bile?" ❌ Overlapping cards → Testing same fact multiple ways ✅ Merge redundant cards; each fact tested once optimally ❌ Recognition vs. recall → Cards too easy (recognition only) ✅ Use cloze deletions and type-in cards for active recall Study Strategy Issues: ❌ Too many new cards → 100+ new cards/day leads to burnout ✅ Sustainable pace: 20-30 new cards/day ❌ No review of missed cards → Ignoring leeches ✅ Tag and review difficult cards separately ❌ Passive card creation → Making cards without studying ✅ Create cards for immediate use; don't build massive backlog References Available in references/ directory: anki_format_spec.md - Anki .apkg file format specifications