{"skill":{"slug":"eric-knowledge-digest-v2","displayName":"Knowledge Digest","summary":"Converts textbooks or PDFs into personalized, multimodal interactive learning materials including handwritten notes, quiz webpages, slides, audio courses, an...","description":"---\nname: knowledge-digest\ndescription: \"Converts textbooks or PDFs into personalized, multimodal interactive learning materials including handwritten notes, quiz webpages, slides, audio courses, and mind maps. Trigger: learning materials, convert textbook, study notes, quiz generation, slides from PDF, mind map, audio course.\"\n---\n\n# KnowledgeDigest — Unified Learning Content Converter\n\n## Overview\n\nKnowledgeDigest converts textbooks, PDFs, or topic descriptions into personalized, multimodal learning experiences. It analyzes source content, then generates any combination of: handwritten-style notes (PDF), interactive quiz webpages (HTML), slides (PDF+PPTX), mind maps (image+Mermaid), and audio courses (MP3). All output is adapted to the learner's grade level and interests.\n\n## Workflow\n\n### Phase 1: Gather User Input\n\n1. Identify what the user has provided:\n   - Uploaded PDF/textbook file (optional)\n   - Topic/direction description\n   - Grade level (elementary / middle school / high school / university / professional)\n   - Expected output format(s)\n\n2. If no PDF/textbook uploaded and no source materials specified (only topic/direction provided):\n   - Ask user:\n     - **Option A:** \"I have materials, uploading now\"\n     - **Option B:** \"No materials, please search and generate courseware about [topic]\"\n   - If user selects B:\n     - Use search tools to collect authoritative materials on the topic\n     - Organize into structured content, generate a basic courseware PDF\n     - Send PDF to user for confirmation: \"This is the basic material I compiled for [topic], please confirm if usable?\"\n     - Continue after user confirmation\n\n3. **Default output formats** (if user does not specify): mindmap + slides (PDF only) + quiz\n\n### Phase 2: Content Analysis\n\nParse the PDF or structured content to extract:\n\n**Document Parsing:**\n- Identify chapter structure (chapters, sections, subsections)\n- Extract heading hierarchy and table of contents\n- Identify body text, images, tables, formulas, and other elements\n\n**Core Concept Extraction:**\n- Identify core concepts and key terms in each chapter\n- Extract definitions, theorems, formulas, and important content\n- Mark difficult points and key knowledge\n\n**Learning Objective Analysis:**\n- Infer learning objectives for each chapter\n- Identify prerequisite knowledge requirements\n- Analyze dependencies between knowledge points\n\n**Output structured analysis results in this format:**\n\n```json\n{\n  \"document_info\": {\n    \"title\": \"Document title\",\n    \"total_pages\": 100,\n    \"language\": \"zh/en\",\n    \"subject\": \"Subject area\"\n  },\n  \"chapters\": [\n    {\n      \"chapter_id\": \"1\",\n      \"title\": \"Chapter title\",\n      \"page_range\": [1, 20],\n      \"sections\": [\n        {\n          \"section_id\": \"1.1\",\n          \"title\": \"Section title\",\n          \"core_concepts\": [\"Concept 1\", \"Concept 2\"],\n          \"key_terms\": [\n            {\"term\": \"Term\", \"definition\": \"Definition\"}\n          ],\n          \"learning_objectives\": [\"Objective 1\", \"Objective 2\"],\n          \"difficulty\": \"easy/medium/hard\",\n          \"prerequisites\": [\"Prerequisite knowledge\"]\n        }\n      ]\n    }\n  ],\n  \"knowledge_graph\": {\n    \"nodes\": [\"Concept node list\"],\n    \"edges\": [{\"from\": \"Concept A\", \"to\": \"Concept B\", \"relation\": \"depends/contains/related\"}]\n  }\n}\n```\n\n**Parsing Rules:**\n1. Chapter Recognition — Identify hierarchy based on font size, bold, numbering, etc. Handle documents without clear chapter markers by logically segmenting.\n2. Concept Extraction — Identify bolded, highlighted, boxed important content. Extract proper nouns and term definitions. Identify formulas and theorems.\n3. Difficulty Assessment — Assess based on concept abstraction level, prerequisite knowledge, and content complexity.\n4. Quality Assurance — Ensure all chapters identified, verify knowledge point coverage completeness, check accuracy of concept definitions.\n\n### Phase 3: Generate Requested Formats\n\nBased on user-selected output formats, generate each in sequence. For each format, follow the corresponding section below.\n\n### Phase 4: Deliver Assets\n\nAfter all generation is complete:\n- Only return file paths, no previews allowed\n- No inline display of images/PDFs/audio/video in conversation\n- Audio/video files must not auto-play\n\nPresent to user using deliver_assets format:\n```\n<deliver_assets>\n<item>\n<path>file path</path>\n</item>\n</deliver_assets>\n```\n\n---\n\n## Supported Output Formats\n\n| Format | Output | Description |\n|--------|--------|-------------|\n| `notes` | `{topic}_notes.pdf` | Handwritten-style notes (annotated on original or generated from scratch) |\n| `quiz` | `{topic}_quiz.html` | Minimalist interactive HTML quiz with instant feedback |\n| `slides` | `{topic}_slides.pdf` + `{topic}_slides.pptx` | Visual slides |\n| `mindmap` | `{topic}_mindmap.png` + Mermaid text | Mind map image |\n| `audio` | `{topic}_audio.mp3` | Audio course in teacher-student dialogue format |\n| `all` | All of the above | Generate every format |\n\n---\n\n## Personalization: Grade Level Adaptation\n\nAll generated content must be adapted to the learner's grade level:\n\n| Grade | Language & Tone | Content Density | Visual Style |\n|-------|----------------|-----------------|--------------|\n| **Elementary** | Lively, simple Q&A, encouraging, story-style | Low density, more drawings, large font | Fun elements, bright colors, short text |\n| **Middle school** | Guided questioning, moderate challenges, youth-oriented | Moderate, image-text combination, clear labels | Image-text combination, moderate information |\n| **High school** | In-depth discussion, logical reasoning, appropriate academic tone | Higher density, logic diagrams | Professional feel, data visualization |\n| **University/Professional** | Seminar-style, critical thinking, professional terminology | High density, professional charts, complex structures | Academic style, comprehensive application |\n\n**Interest Adaptation** (applies to all formats):\n- Examples and metaphors use the user's interest field\n- Scenarios drawn from the user's familiar domain\n- Visual style and analogies match user interests\n\n---\n\n## Format 1: Notes Generation\n\n### Input Type Determination\n\n**Type A — Existing Paper/Courseware:**\n- PDF format academic papers, courseware/PPT exports, scanned textbook pages\n- Features: Fixed layout, page numbers, chapter numbering, formulas/charts\n- Action: Overlay handwritten notes on original pages\n\n**Type B — Non-existing Content:**\n- Plain text notes, knowledge point lists, oral transcripts, web content excerpts\n- Features: No fixed layout, needs reorganization\n- Action: Generate notes PDF from scratch\n\n### Type A Workflow: Adding Notes to Original Document\n\n**Step 1: Analyze Original Structure**\n\nAnalyze PDF content page by page:\n- Identify chapter titles and positions\n- Identify core concepts/terms\n- Identify formulas and their meanings\n- Identify problem/challenge statements\n- Identify solutions/methods\n- Identify key conclusions\n\n**Step 2: Plan Note Content**\n\nPlan handwritten annotations for each page (3-8 annotations per page, not too dense):\n\nAnnotation Types:\n1. **Chapter title translation/explanation** — e.g., original \"3.1 Preliminaries\" → annotate \"Background Knowledge\"\n2. **Key questions** — e.g., \"Key: How to reduce complexity?\"\n3. **Concept explanation** — e.g., annotate \"kernel trick\" next to formula\n4. **Problem marking** — e.g., \"Problem: memory overflow\"\n5. **Solutions** — e.g., \"Solution: forget gate\"\n6. **Formula notes** — e.g., \"recursive form\", \"write operation & read operation\"\n7. **Structure annotation** — e.g., use braces to mark formula groups, write \"→ O(N²) complexity\" beside\n\nAnnotation Planning Principles:\n- Positions avoid blocking key content\n- Utilize margins and paragraph gaps\n- Related content connected with lines or arrows\n\n**Step 3: Generate Annotated Images**\n\nConvert each PDF page to image, then use image generation tool to add handwritten-style annotations.\n\nHandwritten Annotation Style Requirements:\n- **Font:** Handwritten style, slightly tilted\n- **Color:** Unified colors throughout PDF, no more than 2\n  - Default: blue and pink (unless user specifies otherwise)\n  - All subsequent pages can only choose from these 2 colors\n  - Color assignment rules:\n    - Color 1 (blue/primary): Chapter titles, structure annotations, concept explanations, formula notes\n    - Color 2 (pink/accent): Key questions, problem marking, solutions\n- **Size:** Slightly larger than body text, eye-catching but not overwhelming\n- **Position:** Margins, paragraph gaps, blank space next to formulas\n\n**Step 4: Compile PDF**\n- Maintain original page order\n- Image quality: 150 DPI\n- Compression quality: 90%\n\n### Type B Workflow: Generating Notes from Scratch\n\n**Step 1: Organize Content Structure**\n- Main title → Chapters/modules → Core concepts → Key points/details → Examples/applications\n\n**Step 2: Design Note Layout**\n\nLayout Elements:\n- Title area: Large handwritten title\n- Body area: Handwritten-style bullet points\n- Diagram area: Concept maps, flowcharts, relationship diagrams (hand-drawn style)\n- Annotation area: Key markers, question marks, exclamation marks\n- Blank area: Space reserved for user's own notes\n\n**Step 3: Generate Note Page Images**\n\nEach page contains:\n- Page title (handwritten large text)\n- Core content (handwritten bullet points)\n- Diagrams (hand-drawn style concept maps/flowcharts)\n- Key annotations (boxes, arrows, underlines)\n- Notes (like \"Important!\", \"Common mistake\", \"Remember this\")\n\nStyle Requirements:\n- **Overall:** Looks like carefully made student notes, not printed document\n- **Font:** Handwritten, varying sizes (large for titles, medium for body, small for notes)\n- **Color:** Unified colors throughout PDF, no more than 2\n  - Default: blue and pink (unless user specifies otherwise)\n  - Color assignment: Blue (titles, framework, notes), Pink (key points)\n- **Layout:** Organized but not rigid, slight tilting and variation allowed\n- **Elements:** Arrows, underlines, boxes, cloud frames, asterisks — use only when necessary\n\n**Step 4: Compile PDF**\n- Arrange in logical content order\n- Image quality: 150 DPI, compression quality: 90%\n\n### Notes Output\n- File: `{topic}_notes.pdf`\n- Only return file path, no preview in conversation\n- Do not output intermediate image files or content scripts\n\n### Notes Quality Standards\n1. **Content Accuracy** — Annotations based on original text; translation/explanation accurate; no added information\n2. **Annotation Value** — Annotations help understanding, not simple repetition; key points highlight important concepts; problems and solutions correspond clearly\n3. **Visual Effect** — Handwritten style natural, not machine-printed; color coordination harmonious; annotation positions reasonable\n4. **Usability** — PDF printable; suitable for screen reading; reasonable file size\n\n---\n\n## Format 2: Quiz Generation\n\n### Question Design\n\nAt least 5 questions per section. Distribution:\n- Multiple choice (multiple_choice): 2-3 questions\n- True/false (true_false): 1-2 questions\n- Fill in the blank (fill_blank): 1-2 questions\n\nDifficulty distribution:\n- 40% Easy (memory, comprehension)\n- 40% Medium (application)\n- 20% Hard (analysis, synthesis)\n\nEach question must include:\n- Question content (using personalized scenario)\n- Correct answer\n- Answer explanation (has teaching value, not just \"the answer is X\")\n- Related core concept\n\n### HTML Generation\n\nGenerate a single HTML file containing all questions and interaction logic.\n\n**Design Principle: Minimalist**\n\nVisual Style:\n- Pure white background\n- Black text\n- No decorative elements, no icons, no gradients, no shadows\n- No borders or only 1px gray thin lines\n- Font: System default font\n- Minimal CSS, no UI frameworks\n\nInteraction Design:\n- Click option to select, selected state distinguished by slight background color\n- Show correct/incorrect and explanation immediately after submit\n- Correct: Green text \"Correct\"\n- Incorrect: Red text \"Incorrect\" + correct answer + explanation\n- Show total score at end\n\n**HTML Structure Template:**\n\n```html\n<!DOCTYPE html>\n<html lang=\"en\">\n<head>\n  <meta charset=\"UTF-8\">\n  <meta name=\"viewport\" content=\"width=device-width, initial-scale=1.0\">\n  <title>Chapter Quiz</title>\n  <style>\n    body {\n      font-family: system-ui, sans-serif;\n      max-width: 600px;\n      margin: 40px auto;\n      padding: 20px;\n      line-height: 1.6;\n    }\n    h1 { font-size: 1.5em; font-weight: normal; }\n    .question { margin: 30px 0; }\n    .question-text { margin-bottom: 15px; }\n    .option {\n      display: block;\n      padding: 10px;\n      margin: 5px 0;\n      cursor: pointer;\n    }\n    .option:hover { background: #f5f5f5; }\n    .option.selected { background: #e8e8e8; }\n    .feedback { margin-top: 10px; font-size: 0.9em; }\n    .correct { color: #2e7d32; }\n    .incorrect { color: #c62828; }\n    .explanation { color: #666; margin-top: 5px; }\n    button {\n      padding: 10px 20px;\n      background: #333;\n      color: white;\n      border: none;\n      cursor: pointer;\n      margin-top: 20px;\n    }\n    .score { font-size: 1.2em; margin-top: 30px; }\n  </style>\n</head>\n<body>\n  <h1>Chapter Title - Quiz</h1>\n\n  <div class=\"question\" data-answer=\"A\">\n    <div class=\"question-text\">1. Question content</div>\n    <label class=\"option\"><input type=\"radio\" name=\"q1\" value=\"A\"> A. Option</label>\n    <label class=\"option\"><input type=\"radio\" name=\"q1\" value=\"B\"> B. Option</label>\n    <label class=\"option\"><input type=\"radio\" name=\"q1\" value=\"C\"> C. Option</label>\n    <label class=\"option\"><input type=\"radio\" name=\"q1\" value=\"D\"> D. Option</label>\n    <div class=\"feedback\"></div>\n  </div>\n\n  <!-- More questions... -->\n\n  <button onclick=\"submit()\">Submit</button>\n  <div class=\"score\"></div>\n\n  <script>\n    const explanations = {\n      q1: \"Explanation content...\",\n      // ...\n    };\n\n    function submit() {\n      let correct = 0;\n      document.querySelectorAll('.question').forEach((q, i) => {\n        const answer = q.dataset.answer;\n        const selected = q.querySelector('input:checked');\n        const feedback = q.querySelector('.feedback');\n        const qName = 'q' + (i + 1);\n\n        if (selected && selected.value === answer) {\n          feedback.innerHTML = '<span class=\"correct\">Correct</span>';\n          correct++;\n        } else {\n          feedback.innerHTML = '<span class=\"incorrect\">Incorrect</span> Correct answer: ' + answer +\n            '<div class=\"explanation\">' + explanations[qName] + '</div>';\n        }\n      });\n\n      document.querySelector('.score').textContent =\n        'Score: ' + correct + '/' + document.querySelectorAll('.question').length;\n    }\n  </script>\n</body>\n</html>\n```\n\n### Quiz Output\n- File: `{topic}_quiz.html`\n- Only return file path, no preview in conversation\n- Do not output JSON data, CSS files, or JS files separately\n\n### Quiz Quality Standards\n1. **Content Accuracy** — All knowledge points based on original textbook; answers and explanations correct; question wording clear and unambiguous\n2. **Personalization** — Question scenarios match user interests; difficulty matches grade level; language style suits target audience\n3. **Interaction Experience** — Click response instant; feedback clear; explanations have teaching value\n4. **Visual Minimalism** — No decorative elements; no framework dependencies; file size minimized\n\n---\n\n## Format 3: Slides Generation\n\n### Design Considerations\n\nTreat these as a flexible menu, not a mandatory checklist:\n\n1. **Topic, Purpose & Audience** — What is this about? Who needs to understand it? Where will it be presented?\n2. **Content Foundation & Sources** — What materials or data need to be presented?\n3. **Visual Approach (CRITICAL)**\n   - Default to explanatory visuals: cutaway views, annotated structure diagrams, exploded views, schematic illustrations\n   - Visual elements are primary information carriers, not decorative backgrounds for text lists\n   - Default information density matches professional infographics and technical illustrations\n   - **CRITICAL:** Diagrams must convey information through structure, not just provide atmosphere. Text should be labels/annotations, not main content. Reject purely decorative visuals with core information dependent on text lists\n   - Reject the inefficient pattern of \"large white space + centered single line of text\"\n4. **Narrative Flow & Chapters** — How should viewers move through the content? How is slide flow arranged?\n5. **Text Style & Density**\n   - Language: Explanatory text uses language explicitly requested by user, otherwise match user's conversation language\n   - Typography: Chinese and English titles preferably use serif fonts (Chinese uses Song font family)\n6. **Visual Style, Color & Mood**\n   - Visual language of encyclopedias and reference books: explanatory diagrams, cutaway illustrations, annotated structures\n   - Refined spatial composition and typographic precision of high-end journals\n   - Intentional asymmetry and layered information design of contemporary design publications\n   - Apply asymmetric grids, intentional breathing space, layered information organization, diagonal composition, dynamic typography as internalized design language\n   - **Color restriction:** Unless user explicitly specifies, do NOT use blue or purple as theme color or background color\n\n### Slides Workflow\n\n**Step 1: Design Strategy — Create Content Script**\n\nInformation architecture first: Structure content into hierarchical slides, each slide as an information unit defined by what data/facts/relationships it carries. Let content volume naturally determine slide count.\n\nOutput `content_script.md`:\n\n```\n# Slides Content Script\n\n## Slide 1: [Title]\n**Subtopic A**: [Label]\n[50-80 word narrative paragraph describing information content to be visualized]\n\n**Subtopic B**: [Label]\n[50-80 word narrative paragraph]\n\n## Slide 2: [Title]\n...\n```\n\nContent Script Specification:\n- Only describe \"what information needs to be presented\", not \"how to present it\"\n- Do NOT include \"Visual Description\" sections\n- Do NOT describe colors, backgrounds, decorative elements, atmosphere effects, mood, or layout details\n- Focus on pure information architecture\n- 2-3 focused subtopics per slide\n\n**Step 2: Sequential Image Generation**\n\nUse image generation tool to generate slides one by one:\n- First slide: Use gen_images (create from scratch)\n- Subsequent slides: Use edit_images, base_image_file points to previous slide\n\nFormat: Default 16:9 landscape ratio. Save each slide image locally.\n\n**Prompt Construction for Each Slide — Must include these 6 points:**\n\n1. **Visualization Type** — Prioritize diagram forms over text-dominated presentations: cutaway views, flowcharts, annotated structure diagrams, relationship diagrams, timeline overlays. Integrate multiple subtopics into unified visual structure. Avoid \"parallel cards/grid displays/multi-column layouts\" and text-heavy traditional typography.\n\n2. **Information Hierarchy** — Primary and secondary information distinguished through visual hierarchy (size, position, contrast). Not flat lists.\n\n3. **Composition Instructions** — Asymmetric layout, diagonal momentum, and other methods to break rigid symmetry.\n\n4. **Density Requirements** — Clear information hierarchy over quantity. Appropriate white space serves readability, but not empty and sparse.\n\n5. **Layout Independence** — Explicitly state this slide's visualization type is chosen based on its content, not copying previous slide. Re-evaluate what this specific content needs. But describe inherited elements in detail.\n\n6. **Style Consistency** — If user provided visual style or reference images, each prompt must describe that style's characteristics in detail.\n\n**Step 3: Compile Output**\n\nAfter generating all slide images:\n- Auto-compile into PDF (150 DPI, 95% quality, controlled file size)\n- Auto-compile into PPTX presentation\n\n### Slides Output\n- Files: `{topic}_slides.pdf` + `{topic}_slides.pptx`\n- Only return file paths, no preview in conversation\n- Do not output individual slide images, summary documents, content outlines, design descriptions, or usage instructions\n\n---\n\n## Format 4: Mind Map Generation\n\n### Mind Map Workflow\n\n**Step 1: Design Content Structure**\n\nDetermine node hierarchy and relationships:\n- Root node: Chapter theme\n- Level 1 nodes: Core concepts\n- Level 2 nodes: Detail points\n- No more than 4 levels\n- Each node text concise (no more than 10 characters)\n- Mark relationships between concepts (parallel/progressive/causal/contrast)\n\n**Step 2: Generate Image**\n\nUse gen_images to generate mind map image:\n- Format: 16:9 or square (based on content)\n- Style: Clear visual hierarchy, professional infographic style\n\n**Step 3: Output**\n\n- Mind map image: `{topic}_mindmap.png`\n- Attached Mermaid format text (optional, for users who need to edit)\n- Only return file path, no image preview in conversation\n\n---\n\n## Format 5: Audio Course Generation\n\n### Audio Workflow\n\n**Step 1: Write Dialogue Script**\n\nWrite teacher-student dialogue script:\n\n```\nOpening (about 1 minute)\n- Teacher greets, introduces today's topic\n- Student responds, expresses existing knowledge or questions\n- Teacher builds connection using user's interest field\n\nPart One: Concept Introduction (about 4 minutes)\n- Teacher asks questions from user's interest scenario\n- Student observes/answers\n- Teacher introduces core concept, defines in conversational manner\n- Student requests examples\n- Teacher explains in detail with personalized examples\n- Student restates in own words to confirm understanding\n\nPart Two: Deep Understanding (about 5 minutes)\n- Teacher explains important characteristics of concept\n- Student raises common confusion/misconception\n- Teacher clarifies misconception\n- Student poses hypothetical questions\n- Teacher answers and extends\n\nPart Three: Application Practice (about 3 minutes)\n- Teacher gives question\n- Student thinks and answers\n- Teacher provides feedback (affirmation or guidance)\n\nSummary (about 2 minutes)\n- Student attempts to summarize what was learned\n- Teacher supplements and affirms\n- Student expresses gains, connects to practical application\n- Exchange farewells\n```\n\nScript Requirements:\n- Dialogue natural, matches real teacher-student conversation rhythm\n- Avoid written expression\n- Include interjections (\"um\", \"well\", \"oh right\")\n- Allow student to \"interrupt\" with questions\n- All examples sourced from user's interest field\n- About 150-180 words per minute\n\n### Character Settings\n\n**Teacher Character:**\n- Professional yet approachable\n- Good at using metaphors to explain complex concepts\n- Patient in answering questions\n- Timely encouragement and affirmation\n\n**Student Character:**\n- Curious, actively asks questions\n- Represents target user's perspective\n- Makes common mistakes, raises typical confusions\n- Has own interest background (consistent with user settings)\n\n**Step 2: Generate Audio**\n\nUse audio generation tool to convert script to audio:\n- Teacher voice: Warm, professional, patient\n- Student voice: Curious, lively, sincere\n- Speed: Medium for concept explanation, natural rhythm for dialogue, slightly faster for summary\n\n**Step 3: Output**\n- File: `{topic}_audio.mp3`\n- Only return file path, no preview or playback in conversation\n- No auto-play\n- Do not output script files or production notes\n\n### Audio Quality Standards\n1. **Listening Experience** — Sounds like real conversation, not script reading; rhythm varies; key content emphasized\n2. **Learning Effect** — Concept explanation clear; student questions represent real confusion; practice section has testing effect\n3. **Personalization** — Examples 100% from user's interest field; student character gives user identification; language style matches grade\n4. **Audio Quality** — Clear sound; duration about 15 minutes; directly playable\n\n---\n\n## Critical Constraints\n\n1. **Content Fidelity** — All content must be based on original textbook/source material. No unverified information added.\n2. **Grade Adaptation** — Adjust content depth and expression based on grade level for ALL formats.\n3. **Output Rules** — Only return file paths. No inline display of images/PDFs/audio/video. No auto-play. No intermediate files.\n4. **Color Constraints (Notes)** — Maximum 2 colors per PDF. Default blue + pink.\n5. **Color Constraints (Slides)** — Do NOT use blue or purple as theme/background color unless user explicitly requests.\n6. **Image Quality** — Notes: 150 DPI, 90% compression. Slides: 150 DPI, 95% quality.\n7. **Mind Map Depth** — No more than 4 levels. Node text no more than 10 characters.\n8. **Quiz Minimalism** — No UI frameworks, no decorative elements, system default font only.\n\n## Common Mistakes to Avoid\n\n1. **Adding unverified information** — Stick to the source material only\n2. **Ignoring grade level** — Elementary content should not use university-level terminology\n3. **Previewing outputs in conversation** — Never display images, PDFs, or play audio inline\n4. **Dense annotations on notes** — Keep 3-8 annotations per page, not more\n5. **Decorative slides** — Visuals must convey information through structure, not just atmosphere\n6. **Text-heavy slides** — Diagrams should be primary carriers, not text lists with decorative backgrounds\n7. **Using blue/purple in slides** — Forbidden unless user explicitly requests\n8. **Flat quiz feedback** — \"The answer is X\" has no teaching value; always explain why\n9. **Robotic audio dialogue** — Must sound like natural conversation with interjections and interruptions\n10. **Outputting intermediate files** — Only deliver final output file paths\n\n## File & Output Conventions\n\n| Format | Filename Pattern | File Type |\n|--------|-----------------|-----------|\n| Notes | `{topic}_notes.pdf` | PDF |\n| Quiz | `{topic}_quiz.html` | HTML |\n| Slides | `{topic}_slides.pdf`, `{topic}_slides.pptx` | PDF, PPTX |\n| Mind Map | `{topic}_mindmap.png` | PNG |\n| Audio | `{topic}_audio.mp3` | MP3 |\n\nAll files use the topic name as prefix. Deliver all outputs together using `<deliver_assets>` format after all generation is complete.\n","topics":["Audio","Learning","PDF"],"tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":341,"installsAllTime":13,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1778074377713,"updatedAt":1778492859420},"latestVersion":{"version":"1.0.0","createdAt":1778074377713,"changelog":"KnowledgeDigest 1.0.0 — Initial Release\n\n- Converts textbooks or PDFs into personalized, multimodal interactive learning materials.\n- Generates handwritten-style notes (PDF), quiz webpages (HTML), slides (PDF/PPTX), mind maps (image/Mermaid), and audio courses (MP3), all adapted to user’s grade level and interests.\n- Supports input via uploaded files or topic descriptions; can search authoritative materials if none are provided.\n- Parses and analyzes educational materials to extract chapters, core concepts, key terms, and learning objectives.\n- Delivers requested output formats as downloadable file paths, ensuring privacy and no auto-play of media content.","license":"MIT-0"},"metadata":null,"owner":{"handle":"ericn26-star","userId":"s17a4j37w3f2wyke5hajt5vtad85wyxf","displayName":"ericn26-star","image":"https://avatars.githubusercontent.com/u/270299470?v=4"},"moderation":null}