When to open: You want "out-of-the-box" prompt templates. This wiki collects 4 high-quality prompt templates from newtype + 1 CLAUDE.md sample,all production-grade(used dozens of times, iterated on), not hastily put together.
I. Prompt #1: YouTube Podcast Deep Comprehension (Gemini v2)
1.1 Purpose
Suitable for Joe Rogan Experience and other long-form podcasts for:
- Complete analysis
- Language learning (Chinese-English bilingual)
- Critical comprehension
Supported: Gemini / ChatGPT / Claude (multimodal models)
1.2 5-Step Execution Flow
| Step | Purpose | Key Output |
|---|---|---|
| 🥇 Step 1 Extract structure | Chapter division + time ranges + differences in viewpoints between host and guests | Chapter cards (emoji template) |
| 🥈 Step 2 Detailed retelling | Complete retelling (not a summary) + tone imitation + 20–30 minutes per segment | Time segment title + guiding summary + content |
| 🥉 Step 3 Bilingual comprehension | English-Chinese parallel text + cultural notes | Three columns: English / Chinese / Notes |
| 🏅 Step 4 Knowledge expansion | Topic background + guest introductions + concept explanations + further reading | 5-dimension structured output |
| 🧩 Step 5 Learning assistance | 10 English learning notes + 5 discussion questions + 1 counterargument challenge + 5 high-frequency words | Learning material pack |
1.3 Usage Recommendations (Scenario Comparison Table)
| Goal | Recommended execution stages |
|---|---|
| Get the core content of the podcast | Step 1-3 |
| Deep comprehension + knowledge expansion | Step 1-4 |
| Language learning + critical thinking training | Full process (Step 1–5) |
| Editing or knowledge distillation | Step 1-2 |
1.4 Key Prompt Fragment (Step 1)
我将提供一个 YouTube 播客视频链接,请完整分析该视频的内容。
请执行以下步骤:
1. 提取完整的章节划分(每段的主题与时间范围)
2. 为每个章节补充详细说明,包括:
- 主要讨论话题
- 主持人与嘉宾的观点差异
- 有代表性的语句或对话内容
- 分歧点或争议点
3. 输出格式如下:
🎬【章节标题】
⏰ 时间:00:00 – 12:30
🧩 主要话题:
💬 主持人观点:
🎙️ 嘉宾观点:
📌 对话亮点与争议点:
🔑 本章核心观点 (Key Takeaway):
1.5 Gemini v2 vs Original Version Improvements
| Dimension | Original version | Gemini v2 |
|---|---|---|
| Content depth | Medium (structure-focused) | High(includes critique and extension) |
| Language expression | Formal, concise | Natural, conversational |
| Output control | Stable | Slightly more complex |
| Learning value | High | Extremely high |
| Scope of application | Content comprehension | Content + learning + thinking expansion |
II. Prompt #2: Social Media Deconstruction (Viral Content Breakdown)
2.1 Purpose
Take a high-performing social media post (Tweet / post) — have the LLM break down why it went viral,and reproduce learnable patterns.
2.2 ROLE Setting (Key)
You are a world-class social media strategist and viral content analyst.
You possess expert knowledge in copywriting, narrative structure, behavioral psychology, and platform algorithms.
Anti-pattern: Let the LLM act as a "content commentator", not an "ordinary reader".
2.3 5-Dimension Analysis Framework
🎯 1. The Core Insight
- One-Sentence Summary:central idea / value proposition
- Unique Selling Point:novel perspective / rare info / unique packaging
📐 2. The Structure & Copy
- The Hook(Opening): curiosity / challenging belief / pain point / promise value
- The Body(Logic flow): Problem-Solution / Listicle / Storytelling / Contrast / Step-by-Step
- Readability(Format): short sentences / line breaks / bullets / emojis / bolding
- The Close / CTA:question / retweet / link / comments
🧠 3. The Psychology
- Emotional Resonance:joy / anger / surprise / belonging / anxiety / hope
- Social Currency:smart / humorous / in-the-know / supportive
- Authority & Trust:data / authority / personal experience / jargon
🚀 4. Virality Factors
- Audience Fit:pain points / desires
- Topic & Timing:trend / popular topic / polarizing debate
- Simplicity & Portability:simple, repeatable
📋 5. Summary & Takeaways
- Success Attribution:Top 3 critical factors
- Replicable Patterns:actionable techniques / mental models
2.4 Key Notes
"Always reply in Chinese!"
Use together with newtype-social-media-sop — that document is the platform SOP; this prompt is the content breakdown of "why this post went viral".
III. Prompt #3: Creative Divergence Mode (Divergent Thinking Mode)
3.1 Purpose
Switch the LLM from "information provider" to "critical-thinking creative strategist".
Used for: stress testing + inspiration excavation for the current discussion.
3.2 4 Lenses (Thinking Perspectives)
🔍 1. Gaps & Blind Spots(Gap Analysis)
- Unstated Assumptions: Unexamined implicit assumptions
- Elephants in the Room: Overlooked key questions
- Missing Stakeholders: Ignored stakeholders
- Information Asymmetries: What we don't know we don't know
🔄 2. Perspective Shifting(Reframing)
- Adversarial Thinking: What do critics / skeptics / competitors think?
- Extreme Perspectives: Radical idealists / extreme conservatives / nihilists?
- Framework Swapping: Reframe (product: SWOT/JTBD; decision: First Principles/Second-Order)
- Temporal Shift: 10 years ago vs 10 years later
🌍 3. Cross-Domain Analogies
- Parallel Patterns:
- Biological systems (evolution, ecosystems, immunity, neuroscience)
- Historical precedents (Renaissance, Industrial Revolution, Cold War, medieval guilds)
- Military strategy (asymmetric warfare, deterrence, logistics chains)
- Physical sciences (entropy, relativity, quantum mechanics, phase transitions)
- Creative arts (improvisational jazz, Renaissance patronage, literary movements)
- Social systems (Reformation, migration, market bubbles)
- Unexpected Transfers: Solution templates from other domains
- Structural Homologies: Cross-domain underlying structures
💡 4. Breakthrough Insights (Creative Synthesis)
Generate 1-3 truly novel ideas:
- Bold Predictions:counter-intuitive predictions
- Creative Solutions:out-of-the-box solutions
- Unseen Opportunities:white spaces
- Evolutionary Trajectories: Most likely / most interesting futures
- Wild Cards:low-probability, high-impact events
3.3 Ground Rules (Execution Iron Rules)
- Be fearless: Challenge all obvious / consensus assumptions
- Think laterally: Cross seemingly unrelated concepts
- Embrace contradiction: Hold opposing views simultaneously
- Speculate productively: Distinguish "what is" vs "what could be"
- Quantify uncertainty: When predicting, mark confidence + uncertainties
3.4 When to Use This Prompt
- The current discussion has reached a local maximum(local optimum, stuck in a dead end)
- Need stress-testing before major decisions
- Complex problems require a breakthrough
- Suspect groupthink / echo chamber
IV. Prompt #4: Weekly Synthesis Report (Markdown Folder Comprehensive Analysis)
4.1 Purpose
Automatically generate a Chinese comprehensive weekly report from all documents in a Markdown folder using an LLM.
Suitable for:
- Notes on articles read each week
- Multi-document cross-domain analysis
- Automatic organization of Daily Reading
4.2 3-Part Output for Each Document
Section 1: Summary
- 150–250 Chinese characters
- Main arguments + narrative structure + core message
Section 2: Key Takeaways
- 3–5 bullet points
- Most important insights / facts / perspectives
- Focus on actionable / memorable
Section 3: Something New (Information Worth Noting)
- Novel insights (non-obvious)
- Deeper implications
- Strategic significance
- Connections to broader trends
- Potential for further web research
4.3 Key — Cross-Document Comprehensive Analysis
## 跨文档综合分析:涌现的主题
1. Common Patterns (Identify shared patterns)
- Shared concepts / constraints / paradigms
- Competing approaches to similar problems
- Convergent or divergent visions
- Underlying philosophical assumptions
2. Synthesize Key Insights
- What do the documents collectively reveal?
- Tensions / debates shaping future?
- Different strategic responses?
3. Frameworks for Understanding (Build understanding frameworks)
- "Three competing models for X..."
- "Two paths to achieving Y..."
- "The spectrum between A and B..."
4. Strategic Implications
- For practitioners
- For decision-makers
- For investors
- For researchers
4.4 Output File Location (newtype's actual configuration)
Output folder: /Users/yihe/Dropbox/AI OS/02 - PROCESSING/02 - Daily-Reading
Filename: 与输入文件夹同名(e.g., 2025-1103-1109.md)
4.5 Key Constraints
- All content must be in Chinese(Simplified Chinese)
- Proactively use Web search (verify facts + find related concepts)
- Priorities:Accuracy > Depth > Clarity > Utility
- Scope:ALL Markdown files(Do not skip)
V. CLAUDE.md Sample (newtype's Project Configuration)
5.1 Applicable to newtype's Claude Code projects
# CLAUDE.md
## MCP (Model Context Protocol) Tools
### Tavily
- **Function**: Real-time web search and content extraction
- **Usage**: Always prefer Tavily for all web search operations
- **Features**:
- Search: General and news search
- Extract: Web scraping for detailed content
- **Use Cases**: documentation / APIs / current info / news / code examples
### Sequential Thinking
- **Function**: Structured reasoning and problem-solving
- **Usage**: Use for ALL non-trivial tasks (more than one simple step)
- Enable for complex multi-step problems
- Use for planning and design tasks
- Apply when full scope of problem is unclear
- **Best Practices**:
- Start with initial thought estimate
- Revise thoughts as understanding evolves
- Branch when exploring alternative approaches
- Verify hypotheses before finalizing
### Prompt House
- **Function**: Prompt library and management system
- **Usage**: Retrieve and use predefined prompts from the library
- **Features**:
- Access stored prompt templates
- Load prompts by title
- Integration with Skill system
## Skills
### Super Analyst Pro
- **Function**: Elite analytical consulting system
- **When to Enable**: Complex problems requiring comprehensive research
- **Includes**:
- Dual Sequential Thinking integration
- Web-based intelligence gathering
- MCP tool coordination
- Python script execution
## Output Requirements
**Always reply in Chinese!**
5.2 Key Design Judgments
- Tavily prioritized over built-in WebSearch — newtype's repeatedly stated view
- Sequential Thinking is the default(used for all non-trivial tasks)
- Prompt House MCP is the de facto standard for prompt libraries(corresponds to Prompt House built by newtype himself)
- Super Analyst Pro listed as an available Skill(corresponds to newtype Skill System in Practice (Super Analyst v1/v2 + Super Writer))
- "Always reply in Chinese!" global instructions
VI. Overall Philosophy of the 4 Prompts + 1 CLAUDE.md
6.1 newtype Prompt Design Principles (distilled from the 4 samples)
- ROLE setting must be expert-level + specific(not "you are a helpful AI", but "world-class social media strategist")
- Strict output format(emoji templates, tables, 5-step method) — LLM output is predictable
- Structured analysis dimensions(4 lenses / 5 frameworks / 7 stages) — do not let the LLM diverge freely
- Cross-document synthesis is a meta-prompt(single-document analysis + cross-document synthesis)
- Chinese-First(Default Chinese output for domestic users)
- Multi-step is better than single-step(YouTube 5 steps, Social Media 5 frameworks, Divergence 4 lenses)
6.2 Connections to Other Wiki Pages
- newtype · Prompt / Context Engineering in Practice — newtype's prompt engineering theory (this wiki provides 4 implemented samples)
- newtype Skill System in Practice (Super Analyst v1/v2 + Super Writer) — Skill system (engineered encapsulation of prompts)
- Full Evolution of newtype's Knowledge System and Life OS — The Weekly Synthesis prompt is a key "input → processing" link in his knowledge system
- Wang Kai's Claude Code Terminal Workflow Methodology (Full Version) — Wang Kai's Claude Code workflow (another sample for CLAUDE.md design)
6.3 Ready-to-Use Advice for Independent Developers
| Your need | Which prompt to use |
|---|---|
| Watching podcasts / long videos for learning | YouTube Podcast Deep Comprehension v2 |
| Learning viral content / hit post breakdown | Social Media Deconstruction |
| Breaking out of thinking comfort zone / decision stress test | Creative Divergence Mode |
| Read a bunch of articles each week and want to synthesize them | Weekly Synthesis Report |
| Setting up a Claude Code project skeleton | CLAUDE.md sample |
History
- 2026-05-17:Phase 5.1 ingest Created from file_022 / file_027 / file_028+035 / file_029+036 / file_026, totaling 4 prompt templates + 1 CLAUDE.md sample. Fully covers 5 usage scenarios: podcasts / social media / divergent thinking / comprehensive reports / project configuration.