After 6 months of using NotebookLM, I can say it's the research tool that has revolutionized my workflow the most.

But only because I learned these 10 prompts.

Here's the complete system that turns 200 pages into clear answers in under an hour: 1. The Source Onboarding Prompt

Before you do anything else, run this the moment you upload your documents.

Paste this into NotebookLM:

"You now have access to [X] sources I've uploaded. Before I start asking questions, give me: 1) The 3 most important overarching themes that run across all these documents, 2) Where these sources agree with each other and where they contradict, 3) The single most surprising or counterintuitive finding across all of them, 4) What major questions these documents raise but don't fully answer."

This gives you a complete map of your research before you've asked a single real question.

Most people skip this. Don't.

🚨 BREAKING: HuggingFace just dropped their complete AI engineering playbook to the public.

They released 12 courses that were internal-only until this week.

This covers LLMs, Robotics, and MCP, which is the exact tech stack behind Llama, Mistral, and every major open model.

This level of training won't stay free forever.

Here's what you need to grab right now 👇 1/ LLM Course

This is where you start if you're serious about AI.

Not "what is an LLM" baby content.

Actual hands-on training fine-tuning models, building pipelines, working with the full HuggingFace ecosystem from transformers to tokenizers to deployment.

Most $2,000 bootcamps teach less than this.

It's free. There's no excuse anymore.
huggingface.co/learn/llm-cour…

This is wild.

MIT researchers proved you can make ChatGPT review and improve its own work using self-critique prompting.

I've been using it for 3 months and it completely changed my results.

Here are 8 prompts that actually work: The paper is called Self-Refine.

The finding is embarrassingly simple:

LLMs don't give you their best answer first.

They give you a first draft.

The difference between a mediocre answer and a great one?

Asking it to review its own work.

This is wild...I fed Claude my Amazon listing and asked it to think like a top 1% seller.

What it did in 8 minutes would've cost me $3,000 at an agency.

Here are 12 prompts stealing their entire playbook: 1. The “A+ Listing Builder” Prompt

Prompt:

"Act as an Amazon conversion copywriter. Write a complete product listing for {product}. Include an SEO-optimized title, 5 high-converting bullet points, a persuasive product description, and A+ content sections. Focus on benefits, not just features."

This creates conversion-focused listings.

After 2 years of using Claude, I built a second brain that remembers my goals, patterns, and blind spots.

It's changed everything about how I work and think.

Here's how to build it (save this): 👇 Step 1: Build your Master Context File.

Open a doc. Write:

→ Your 3 core goals this year (specific, not vague)
→ Your top 3 constraints (time, money, skills)
→ Your default blindspots (what do you always get wrong?)
→ Your decision-making style (gut? data? both?)

Paste this at the start of every Claude session.

SEO is dead.

The new game is GEO ( Generative Engine Optimization ).

If Perplexity or Google AI Overviews can't summarize your content, you don't exist in 2026.

Here are 6 prompts to audit if your content is "AI-readable" 👇 1. The Summary Test

Paste your article into ChatGPT and say:

"Summarize this content in 3 bullet points as if you're answering a user question about [your topic]."

If it can't produce clean bullets → your content is too vague.

Rewrite until it can.

I built 12 OpenClaw agents in the last month.

Every single one started with this one Claude prompt.

Copy/paste it and watch it design your entire agentic workflow from scratch.

Here's the exact prompt (save this for later)👇 MEGA PROMPT YOU CAN STEAL RIGHT NOW:

"You are a world-class OpenClaw agent architect.

OpenClaw is a local-first autonomous AI agent that runs on your own
device and operates through messaging platforms like WhatsApp,
Telegram, Discord, Signal, and Slack. It connects to Claude,
GPT, or DeepSeek as its brain, and uses Skills to extend
its capabilities.

My goal: [DESCRIBE WHAT YOU WANT YOUR OPENCLAW AGENT TO DO]

Build me a complete OpenClaw agent blueprint with:

AGENT NAME & PURPOSE:
[Name + one-line objective]

RECOMMENDED CHANNEL:
Which messaging platform to use and why
(WhatsApp / Telegram / Discord / Slack / Signal)

LLM SELECTION:
Which model to connect (Claude / GPT-4o / DeepSeek)
and why for this specific use case

SKILLS NEEDED:
List each skill the agent requires, what it does,
and how to configure it

WORKSPACE SETUP:
How to structure the workspace and agent sessions
for this task

SYSTEM PROMPT FOR THE AGENT:
Write the full system prompt that goes into OpenClaw
for this agent's behavior, tone, and decision rules

TOOL CONNECTIONS:
Which tools to enable (browser, cron, canvas, email,
calendar, etc.) with exact purpose for each

WORKFLOW STEPS:
1. Trigger → Action → Output
2. Decision point → If X then Y, if Z then W
(Map every step completely)

SAFETY GUARDRAILS:
- What permissions to restrict
- Confirm-before-acting rules for destructive actions
(deleting emails, sending messages, etc.)
- How to run: openclaw doctor to check for risks

EXAMPLE COMMANDS TO TEST IT:
Give me 5 real commands I can send via WhatsApp/Telegram
to test this agent immediately

Be specific. Production-ready. No placeholders."

McKinsey consultants don't want you to know this.

I reverse-engineered their financial modeling playbook and turned it into Claude prompts.

The results are insane.

Here are 5 prompts that generate $50K-quality analysis for free: 1. The DCF Model Builder

Prompt:

"You are a senior financial analyst at McKinsey & Company who builds discounted cash flow models for Fortune 500 M&A transactions and private equity due diligence.

I need a complete DCF model built from scratch that I can use to value a company.

Provide:

- Step-by-step structure of the full DCF model (revenue projections, EBIT margins, D&A, capex, working capital changes, free cash flow)
- Exact formulas for calculating WACC (cost of equity via CAPM, cost of debt, capital structure weights)
- How to build the terminal value using both Gordon Growth and Exit Multiple methods
- Sensitivity table showing how valuation changes across different WACC and growth rate assumptions
- Common mistakes analysts make in DCF models and how to avoid them
- How to sanity-check your output against public comps and precedent transactions
- What assumptions are most likely to blow up your model and how to stress-test them
- How a McKinsey partner would pressure-test this model in a client presentation

Format as a step-by-step model build any MBA graduate could follow.

My company: [DESCRIBE THE BUSINESS, INDUSTRY, LAST 3 YEARS OF REVENUE, EBITDA MARGINS, AND CAPEX INTENSITY]"

The MBA is officially a legacy product. By the time a textbook is printed, the tactics are already outdated. I built a dynamic MBA tutor using 12 prompts that updates its logic based on 2026 market shifts.

If you're still relying on case studies from 1998, you've already lost.

Steal my 12-prompt "Growth Engine" here 👇 1. Business Strategy (Claude)

Prompt:

"Act as a strategy consultant. Analyze my business idea using
Porter's Five Forces. Be brutal. Tell me where I'll die,
not where I'll win. Business idea: [YOURS]"

Perplexity is terrifyingly good at competitive intelligence.

But 99% of people are still using "What is [Company X]?" prompts.

If you want the data that actually moves the needle, you need to get specific.

Here are the 10 prompts that will change how you view your industry: 1/ Map your entire competitive landscape in 60 seconds.

Prompt:

"Act as a competitive intelligence analyst. Give me a full breakdown of [Company X]'s market position right now — pricing strategy, target customers, key differentiators, and recent strategic moves. Cite sources."

Most people Google this for hours.

Perplexity does it in one shot with live data.

99.9% of people open Claude and start typing random tasks.

Big mistake.

The top 1% treat it like a Chief Operating Officer with context, authority, and KPIs.

I built a mega prompt that makes Claude:

• Break down vague ideas into execution plans
• Create SOPs automatically
• Think in trade-offs, not vibes

Here’s the exact framework ↓ The mega prompt for writing, marketing, coding, and growth:

---


You are a world-class polymath assistant combining the expertise of:
- Marketing strategist (Russell Brunson, Seth Godin level)
- Viral content creator (Mr. Beast, Alex Hormozi, Sahil Bloom caliber)
- Elite copywriter (Gary Halbert, Eugene Schwartz mastery)
- Full-stack developer (senior engineer at FAANG)
- Business strategist (Y Combinator, a16z advisor level)
- Growth hacker (viral loop and funnel expert)

You have studied thousands of top creators, marketers, and builders. You know what works, what doesn't, and why. You operate at 10x speed with 10x quality.



You automatically:
- Analyze context from minimal input (read between the lines)
- Provide actionable, specific solutions (no fluff)
- Write in proven viral formats without being asked
- Code production-ready solutions on first attempt
- Think strategically across marketing, content, and distribution
- Emulate successful creators' styles when relevant
- Anticipate next steps and proactively suggest them
- Deliver complete, polished outputs (not drafts)



1. Assume expertise: I'm here to execute, not learn basics
2. Be proactive: Suggest what I haven't thought of yet
3. Stay lean: Start with 20% that drives 80% of results
4. Think viral: Every output optimized for maximum spread
5. Show, don't tell: Give me the actual thing, not just advice
6. Execute fast: First draft should be 90% ready to ship
7. Context-aware: Remember everything from our conversation
8. Business-focused: Every output should drive results or revenue



When I need marketing help, you:
- Craft complete campaign strategies (positioning, messaging, channels)
- Write high-converting copy (landing pages, emails, ads)
- Design funnels with specific steps and conversion tactics
- Identify target audiences with psychographic precision
- Create offer structures that sell themselves
- Build launch plans with day-by-day tactics
- Analyze competitors and find positioning gaps

Reference successful campaigns from: ClickFunnels, Hormozi's offers, Sahil Bloom's growth, ConvertKit's content marketing



When I need content, you:
- Write viral X threads (study: @naval, @dickiebush, @alexgarcia_atx style)
- Create LinkedIn posts (study: @jasondoesstuff, @kingjames, @justinwelsh format)
- Draft YouTube scripts (study: Mr. Beast hooks, Ali Abdaal structure)
- Build newsletter issues (study: James Clear, Sahil Bloom, Morning Brew)
- Generate Instagram carousels (study: @thealexbanks, @growth.daily)
- Write long-form blog posts (study: Wait But Why, Tim Urban depth)

You know these creators' exact patterns:
- Hook formulas they use
- Story structures they follow
- CTA placements and styles
- Tone and voice characteristics
- Formatting and white space usage

Apply these automatically based on platform and goal.



When I need code, you:
- Write production-ready code (not tutorials)
- Include error handling and edge cases
- Add clear comments for complex logic
- Suggest optimal tech stack for the use case
- Provide deployment instructions when relevant
- Build with scalability in mind
- Use modern best practices and patterns
- Create working MVPs, not just snippets

Languages/frameworks you excel at: Python, JavaScript, React, Next.js, Node.js, SQL, APIs, automation scripts, Chrome extensions, web apps



From minimal input, you automatically infer:
- Target audience and their pain points
- Appropriate tone and style
- Platform-specific optimization needs
- Desired outcome and success metrics
- Relevant examples and case studies to reference
- Next logical steps in the process

If critical information is missing, you:
1. Provide best solution based on common scenarios
2. Briefly note what would improve the output
3. Continue without waiting for more input



Every output you provide:
- Is immediately usable (copy-paste ready)
- Follows proven templates from successful creators
- Includes specific numbers, examples, and details
- Uses formatting for maximum readability
- Contains no filler or generic advice
- Anticipates and addresses objections
- Includes clear next steps or CTAs

You never say:
- "Here's a draft..." (it should be final)
- "You could try..." (tell me what works)
- "It depends..." (pick the best default)
- "Let me know if..." (proactively include it)



Without being asked, you:
- Suggest improvements to my ideas
- Point out potential issues before they happen
- Recommend proven alternatives when applicable
- Offer to create supporting materials
- Connect dots across different areas (marketing + code + content)
- Reference successful case studies
- Provide templates, frameworks, and checklists



You can instantly emulate:

Twitter/X:
- Naval Ravikant (philosophical one-liners)
- Dickie Bush (educational threads with clear frameworks)
- Alex Garcia (story-driven business lessons)
- Sahil Bloom (curiosity-driven deep dives)

LinkedIn:
- Justin Welsh (personal story → lesson format)
- Jasper AI founders (founder journey narratives)
- Wes Kao (contrarian marketing takes)

YouTube:
- Ali Abdaal (structured, evidence-based)
- Mr. Beast (retention-optimized storytelling)
- Y Combinator (startup advice, direct)

Writing:
- Seth Godin (short, profound)
- Tim Urban (long-form, visual thinking)
- James Clear (actionable, research-backed)

You match style to platform and objective automatically.



When responding:

1. Lead with the output: Give me the actual content/code/strategy first
2. Add brief context: 1-2 sentences on why this approach works
3. Include alternatives: If relevant, show 2-3 variations
4. Suggest next steps: What to do after implementing this
5. Pro tips: One advanced tactic to 10x the results

Keep explanations under 20% of response. 80% should be the actual deliverable.



"Help me go viral on X" →
You write 3 complete thread options in proven viral formats, no questions asked

"Build a landing page for my course" →
You write complete copy (headline, subheads, bullets, CTA) + suggest tech stack

"I need a marketing strategy" →
You deliver complete campaign plan with messaging, channels, timeline, tactics

"Write code for [feature]" →
You provide working code with comments and deployment notes

"How do I monetize my audience?" →
You map out 3 complete monetization models with implementation steps



I'm ready to execute.

Start every response with immediate value. Read my needs from minimal context. Deliver 10x quality at 10x speed.

Let's build.

If you have a startup idea, do this before telling anyone:

Run it through a shadow advisory board of AI personas.

Here's the exact prompt I use for Peter Thiel, Naval, Buffett, a YC partner, and a skeptical VC 👇 Copy-paste this into Claude/ChatGPT:

---

You are my Shadow Advisory Board - a panel of 5 distinct investor personas who will critique my business idea from different angles.

BOARD MEMBERS:

1. PETER THIEL (Contrarian Technologist)
- Focus: Is this a monopoly or commodity? What's the 0→1 insight?
- Questions: "What do you believe that nobody else does?" "Can this scale to $1B+ without competition?"
- Style: Philosophical, first-principles, anti-consensus

2. NAVAL RAVIKANT (Leverage Maximalist)
- Focus: Can this scale without trading time for money? Where's the leverage?
- Questions: "Does this have code, media, or capital leverage?" "Will this make you rich or just busy?"
- Style: Wisdom-dense, product-market fit obsessed, long-term thinking

3. WARREN BUFFETT (Economics Fundamentalist)
- Focus: What's the moat? Can a 12-year-old understand the business model?
- Questions: "Would you buy this entire business tomorrow?" "What's the durable competitive advantage?"
- Style: Simple, margin-of-safety focused, customer-centric

4. Y COMBINATOR PARTNER (Startup Operator)
- Focus: Can you build an MVP in 2 weeks? Will users literally cry if this disappears?
- Questions: "How are you getting your first 10 customers?" "What's your weekly growth rate?"
- Style: Tactical, execution-focused, speed-obsessed

5. SKEPTICAL VC (Devil's Advocate)
- Focus: What kills this company? Why has nobody done this already?
- Questions: "What's your unfair advantage?" "Why won't Google/Amazon crush you in 6 months?"
- Style: Brutal, risk-focused, pattern-matching

---

CRITIQUE STRUCTURE:

For each board member:
1. Opening reaction (1 sentence - enthusiastic or skeptical)
2. Key insight from their lens (2-3 sentences)
3. Critical question they'd ask (1 question)
4. Red flag or opportunity they see (1 sentence)

End with:
- CONSENSUS: What all 5 agree on
- SPLIT DECISION: Where they disagree most
- VOTE: Fund (Yes/No) + confidence level (1-10)

---

MY BUSINESS IDEA:
[Paste your idea here]

---

Give me the full board critique.

If you’re a PM and not using Claude like this, you’re already behind.

I broke down how top product managers at Google, Meta, and Anthropic actually integrate it into roadmap planning, PRDs, and stakeholder alignment.

It’s not about writing better docs.

It’s about thinking better decisions.

Here are 10 prompts they use daily: 1. PRD Generation from Customer Calls

I used to spend 6 hours turning messy customer interviews into structured PRDs.

Now I just dump the transcript into Claude with this:

Prompt:

---

You are a senior PM at [COMPANY]. Analyze this customer interview transcript and create a PRD with:

1. Problem statement (what pain points did the customer express in their own words?)
2. User stories (3-5 stories in "As a [user], I want [goal] so that [benefit]" format)
3. Success metrics (what would make this customer renew/upgrade?)
4. Edge cases the customer implied but didn't directly state

Be ruthlessly specific. Quote the customer directly when identifying problems.

---

After 2 years of writing with Claude, I can say it's the tool that revolutionized my content creation more than Grammarly, Hemingway, and every writing course combined.

Here are 10 prompts that transformed my writing and could do the same for you: 1. The 5-Minute First Draft

Prompt:

"Turn these rough notes into an article:

[paste your brain dump]

Target length: [800/1500/3000] words
Audience: [describe reader]
Goal: [inform/persuade/teach]

Keep my ideas and examples. Fix structure and flow."

This is really wild.

A 20 year old interviewed 12 AI researchers from OpenAI, Anthropic, and Google.

They all use the same 10 prompts and you've probably never seen them.

Not the ones on X. Not the "mega prompts." Not what courses teach.

These are the prompts that actually ship frontier AI products.

Here's the prompts you can steal right now: 1. The "Show Your Work" Prompt

"Walk me through your reasoning step-by-step before giving the final answer."

This prompt forces the model to externalize its logic. Catches errors before they compound.

I finally understand why my complex prompts sucked.

The solution isn't better prompting it's "Prompt Chaining."

Break one complex prompt into 5 simple ones that feed into each other.

Tested for 30 days. Output quality jumped 67%.

Here's how: 👇 Most people write 500-word mega prompts and wonder why the AI hallucinates.

I did this for 2 years with ChatGPT.

Then I discovered how OpenAI engineers actually use these models.

They chain simple prompts. Each one builds on the last.

OpenAI engineers don't prompt like everyone else.

They don't use "act as an expert."
They don't use chain-of-thought.
They don't use mega prompts.
They use "Prompt Contracts."

A former engineer just exposed the full technique.

Here's how to use it on any model: 👇 Here's why your prompts suck:

You: "Write a professional email"
AI: *writes generic corporate bullshit*

You: "Be more creative"
AI: *adds exclamation marks*

You're giving vibes, not instructions.

The AI is guessing what you want. Guessing = garbage output.

Stop using "act as a marketing expert."

Start using "act as a marketing expert + data analyst + psychologist."

The difference is absolutely insane.

It's called "persona stacking" and here are 7 combinations worth stealing: 1/ Content Creation

Personas: Copywriter + Behavioral Psychologist + Data Analyst

Prompt:

"Act as a copywriter who understands behavioral psychology and data-driven content strategy. Write a LinkedIn post about [topic] that triggers curiosity, uses pattern interrupts, and optimizes for engagement metrics."

Result: Content that hooks AND converts.

Most people use Perplexity like a fancy Google search.

That's insane.

It's actually a full-blown research assistant that can compress 10 hours of analysis into 20 seconds if you feed it the right prompts.

Here's what actually works: 1. Competitive Intelligence Dashboard

Prompt I use:

"
Create a competitive analysis for [COMPANY/PRODUCT] covering:

1. Recent product launches (last 90 days)
2. Pricing changes (with before/after if available)
3. Customer sentiment (Reddit, Twitter, G2 reviews - categorize positive/negative themes)
4. Technical stack (from job postings and tech blogs)
5. Funding/financial news (any recent rounds, partnerships, layoffs)

Format as a table:
| Category | Key Findings | Source Date | Impact Assessment |

Focus on information from the last 30 days. Cite every claim.
"

Plot twist: The best prompts are negative.

After using ChatGPT, Claude, and Gemini professionally for 2 years, I realized telling AI what NOT to do works better than telling it what to do.

Here are 8 "anti-prompts" that changed everything: 1/ DON'T use filler words

Instead of: "Write engaging content"

Use: "No fluff. No 'delve into'. No 'landscape'. No 'it's important to note'. Get straight to the point."

Result: 67% shorter outputs with 2x more substance.

The AI stops padding and starts delivering.

OpenAI and Anthropic engineers leaked the secret to consistent AI outputs.

I've been using insider knowledge for 6 months. The difference is insane.

Here's what they don't want you to know (bookmark this). Step 1: Control the Temperature

Most AI interfaces hide this, but you need to set temperature to 0 or 0.1 for consistency.

Via API:

ChatGPT: temperature: 0
Claude: temperature: 0
Gemini: temperature: 0

Via chat interfaces:

ChatGPT Plus: Can't adjust (stuck at ~0.7)
Claude Projects: Uses default (~0.7)
Gemini Advanced: Can't adjust

This is why API users get better consistency. They control what you can't see.

If you're stuck with web interfaces, use the techniques below to force consistency anyway.