Futuristic cover of The Ultimate Prompt Engineering Bible with AI circuit patterns
Back to White Papers
White Paper

The Ultimate Prompt Engineering Bible

Maverick's ChatGPT Secret Codes — Fully Expounded

100+ Codes • Exhaustive Examples • Research-Backed Techniques • Platform-Specific Optimizations

Published: June 19, 202645 min readBy Maverick MaltinAbout the Author ↓
Maverick Maltin — AI Prompt Engineering Expert & Digital Strategist

Maverick Maltin

Digital Strategist & Prompt Engineer

Table of Contents

How These Codes Work: The Science Behind the Shortcuts

These shortcuts function as high-signal meta-instructions. Large language models are instruction-tuned to follow explicit directives in the context window. Placing a precise code at the end (or integrated) biases token prediction toward specific behaviors: reduced verbosity, structured output, deeper reasoning, or stylistic shifts.

Research in prompt engineering (including chain-of-thought studies by Wei et al. and role/persona effectiveness in instruction-tuned models) shows that explicit format, role, and reasoning instructions dramatically improve output quality, consistency, and depth. Codes like ELI5 leverage simplification training data; critique codes (REDTEAM, KILLCRITIC) activate self-critique patterns seen in AI safety and debate prompting literature.

🎯 Universal rules that amplify every code:

  • Specificity before the code beats vague + code.
  • Stacking codes creates compound effects (e.g., /expand + /analogies + ELICollege).
  • Iteration in conversation often outperforms one-shot prompting.
  • Delimiters (###, quotes, or XML-style tags) help when content might interfere with the instruction.
Abstract visualization of AI neural pathways and token prediction for prompt engineering

AI neural pathways: how prompt codes bias token prediction

Core Writing & Style

/human — Forces natural, conversational prose with varied sentence length, contractions, subtle idioms, and avoidance of repetitive AI patterns ("Furthermore," "It is important to note"). It counters the polished-but-robotic default of many instruction-tuned models.

Example: Rewrite this investor update for Digital Marketing Co. so it reads like a founder wrote it casually over coffee. /human

/rewrite — Rephrases while preserving core meaning. Useful for refreshing content, avoiding detection filters, or shifting perspective.

/improve — Upgrades clarity, flow, rhythm, word choice, and persuasiveness. The universal "make it better" enhancer.

/shorten — Aggressively condenses while retaining every essential point. Excellent for executive summaries or mobile consumption.

/expand — Adds depth, examples, supporting arguments, data, and context. Turns seeds into full sections.

/simplify — Plain language, short sentences, minimal jargon. Ideal for broad audiences or onboarding materials.

/formal — Professional, authoritative tone suitable for proposals, reports, or executive communication.

/casual — Relaxed, conversational tone for social content, newsletters, or community engagement.

💡 Pro tip: Combine /human + /improve for content that sounds genuinely human and persuasive. Learn more about content strategies on our content marketing page.

Holographic floating code blocks representing AI writing and style commands

Writing & style codes: holographic AI command panels

Explanation Levels

These control cognitive load and technical depth.

ELI5Explains as if to a curious child (concrete analogies, zero jargon).
EL10Slightly older child; introduces basic concepts.
EL15Teen level with moderate detail.
ELICollegeUndergraduate rigor and terminology.
ELIPhDDense, precise, assumes domain expertise.

/analogies — Forces comparison-based explanations. One of the highest-leverage techniques for genuine understanding (backed by cognitive science on analogical reasoning).

/deep — Demands layered, multi-dimensional exploration.

/why — Focuses on root causes and mechanisms rather than description.

/firstprinciples — Breaks ideas to fundamental truths then rebuilds. Powerful for innovation and avoiding status-quo thinking.

🔥 Power combo: ELICollege + /analogies + /why — for teaching complex topics like valuation math or neural net training.
Ascending holographic knowledge layers representing AI explanation levels from simple to complex

Explanation levels: from ELI5 to ELIPhD

Formatting & Output

TLDR

Ultra-concise 2-3 sentence summary. Perfect for getting the gist fast.

BULLETS

Formats response as scannable bullet points. Great for notes and meeting recaps.

TABLE

Structures data into Markdown rows/columns. Excellent for side-by-side comparisons.

CHECKLIST

Actionable checklist with checkboxes. Useful for workflows and audits.

STEPS

Numbered step-by-step guide. Perfect for tutorials and procedures.

/outline

Hierarchical outline with headings and nested sub-sections.

/summary

Comprehensive summary covering all main points with balanced weighting.

/keypoints

Extracts only the most important, actionable takeaways.

/timeline

Presents events or steps in chronological order. Great for roadmaps.

/matrix

2×2 or N×N grid for mapping options against criteria.

Explicit format instructions are among the most reliable prompt techniques because they directly constrain output structure, reducing hallucination and improving scannability (widely validated in Microsoft, OpenAI, and academic prompt engineering resources).

📊 Looking for ways to structure your marketing reports? Check out our analytics & reporting services.

Learning & Study

QUIZME

Generates quiz questions

FLASHCARDS

Q&A pairs for spaced repetition

SOCRATIC

Guides via targeted questions

FEYNMAN

Simplest accurate explanation

MISCONCEPTIONS

Surfaces common false beliefs

/teachme

Frames response as structured lesson

Brainstorming & Ideas

IDEAS10

Generates exactly 10 diverse ideas. Forces breadth over depth.

HOOKS10

10 opening hooks for content — headlines, first lines, or email subjects.

ALT3

3 alternative approaches to the same problem. Great for A/B testing.

IDEAGEN

Continuous idea generation mode; keeps producing until you stop it.

SCAMPER

Applies the SCAMPER framework: Substitute, Combine, Adapt, Modify, Put to other use, Eliminate, Reverse.

REMIX

Recombines existing elements in novel, unexpected ways.

/brainstorm100

Mega-session: generates 100 ideas in rapid-fire, prioritizing volume over perfection.

/viralhooks

Hooks engineered for shareability and social media engagement.

/whatif

Hypothetical scenarios to explore unconventional possibilities.

/mashup

Combines two unrelated concepts for cross-domain innovation.

These leverage divergent thinking patterns; quantity prompts often surface more novel options before convergence.

Futuristic mind map with glowing neural connections for AI brainstorming and ideation

AI ideation: divergent neural connections

Analysis & Decision Making

PROSCONS
SWOT
COMPARE
/assumptions
/counter
/debate
/steelman
/devilsadvocate
RISKMAP
SECONDORDER
REDTEAM
DECIDE
PARETO

Techniques like steelmanning, red-teaming, and assumption surfacing draw from AI safety, debate prompting, and decision science literature. They improve robustness and reduce confirmation bias.

🏆 Elite combo: Idea + /steelman + /devilsadvocate + SECONDORDER + RISKMAP + PARETO

🎯 Need strategic analysis for your business? Our free website auditor analyzes 12 dimensions of your digital presence.

Holographic SWOT analysis and decision-making dashboard with glowing AI charts

AI-powered analysis & decision frameworks

Advanced Thinking Systems

These higher-order frameworks push reasoning depth:

DELTR

Decompose, Evaluate, Limit, Transform, Rebuild

OODA

Observe, Orient, Decide, Act

KILLCRITIC

Brutal, no-holds-barred weakness critique

LENSSTACK

Multiple simultaneous perspectives

BLACKSWAN

Focus on rare, high-impact possibilities

REVERSEENGINEER

Work backward from desired outcome

SYSTEMMAP

Actors, incentives, feedback loops

META

Examines the framing of the question itself

Creative & Content

/simulate

Simulates a hypothetical scenario with detailed variables, actors, and outcomes.

/predict

Future projection based on current trends and available data.

/futurehistory

Writes as if reporting from the future about an event that already happened.

/reverse

Reverse engineers: works backward from the outcome to uncover the steps.

SCRIPT

Generates scripts with dialogue, stage directions, and professional formatting.

THREAD

Creates a Twitter/X-style thread optimized for engagement and shareability.

/storymode

Transforms dry information into engaging narrative with dramatic arc.

/persona

Writes from the perspective of a specific character or persona.

/headlines

Generates multiple headlines optimized for clicks and SEO.

/emailcopy

Persuasive email copy with subject line, pre-header, and CTA.

🚀 Discover how creative content drives rankings on our SEO services page.

Coding

DEBUG

Identifies likely bugs, root causes, and suggests fixes.

EXPLAINCODE

Line-by-line plain-language breakdown. Perfect for learning.

OPTIMIZE

Improves speed, readability, and reduces code complexity.

REFACTOR

Restructures code for better maintainability without changing behavior.

UNITTEST

Generates unit tests covering normal, edge, and error cases.

SECURITY

Security audit: scans for vulnerabilities, injections, and data leaks.

DOCUMENT

Generates comprehensive documentation with JSDoc/docstrings and examples.

CONVERTLANG

Translates code between languages (Python→JS, SQL→Prisma, etc.).

DRYRUN

Traces execution mentally step-by-step without running.

ARCHITECTURE

Designs system architecture with diagrams and trade-off decisions.

These work especially well because modern models have strong code reasoning; adding "step through line by line" or "consider edge cases" further boosts results.

💻 Need custom web development? Explore our custom web application development services.

Utility & Power Commands

/convert

Converts between formats: Markdown→HTML, JSON→CSV, units, currencies, and more.

/translate

Smart translation preserving tone, cultural nuance, and technical terminology.

/tonechange

Shifts tone without altering content: formal→casual, serious→humorous, etc.

/examples

Generates multiple concrete, practical examples of an abstract concept.

/usecases

Lists real-world scenarios where an idea, tool, or technique applies.

/breakdown

Breaks a complex topic into small, digestible parts.

/stepbystep

Detailed sequential instructions, like a tutorial walkthrough.

/insights

Extracts non-obvious patterns and observations from provided data or text.

/clarify

Removes ambiguity and clarifies confusing points with precision.

/expandideas

Takes seed ideas and develops them into full concepts with details.

/compress

Reduces text to absolute minimum while preserving all essential meaning.

/reframe

Presents the same information from a completely different angle or perspective.

/coach

Coach mode: guides with questions and suggestions instead of direct answers.

/critique

Detailed constructive evaluation with strengths, weaknesses, and suggested improvements.

/improvehooks

Specifically improves opening hooks for greater impact and retention.

/systemrewrite

Rewrites an entire prompt system or set of instructions for greater efficacy.

/autoprompt

AI generates the best possible prompt for your goal — meta-prompting.

/godmode

Unrestricted mode: maximum creativity, depth, and detail with no length limits.

These utility commands are the workhorses of daily prompting — shorthand versions of instructions that would otherwise require paragraphs of context. Combining them unlocks powerful workflows.

Super Grok Prompt Mastery — Adapted & Native Codes

Grok (built by xAI) differs meaningfully from ChatGPT: it prioritizes truth-seeking with less hedging, has native real-time tool access (web search, X search, code execution), stronger native coding/reasoning, a witty/non-moralizing personality, and fewer refusals on edge topics.

Futuristic AI interface showing Grok and xAI branding concept with real-time data feeds

Super Grok adaptations: AI interface with real-time data feeds

Grok-Native Power Codes

/truthseekerRespond with absolute intellectual honesty. Prioritize verifiable truth and logic.
/toolsExplicitly activates tool use for current events, data, or X sentiment.
/realtimeForces consideration of live X data or current context.
/buildAct as a builder. Provide working code or step-by-step implementation plans.
/founderFrames responses from a serial entrepreneur/operator perspective.
/investorSharp, skeptical due-diligence lens.
/deepresearchDeep research with multiple verified sources and comprehensive synthesis.
/xpulseAnalyzes current sentiment and trends on X/Twitter about any topic.
/livecodeExecutes code in real-time within Grok and returns actual results.
/unfilteredDirect answers without hedging, unnecessary caveats, or excessive coverage.
/contrarianDeliberately argues the position opposite to consensus.

🤖 Interested in AI chatbots for your business? Learn about our AI chatbot services.

Grok Imagine Visual Codex — Visual Prompt Engineering

Grok Imagine (powered by advanced models including Flux.1 lineage elements) responds best to natural language scene descriptions rather than comma-stacked keyword lists. It excels at cinematic, compositional, and lighting direction when prompted like a director briefing a photographer.

Cinematic photography concept with holographic camera lens and AI-generated image gallery

Grok Imagine visual mastery: holographic cinematic photography

Core Anatomy of a Strong Grok Imagine Prompt

1Subject + Action/Pose (lead with this — models weight early tokens heavily).
2Environment / Setting.
3Lighting + Mood / Atmosphere.
4Composition / Camera / Lens / Angle.
5Style / Artistic References.
6Technical Quality (photorealistic, sharp focus, intricate details).

Maltin, Maverick. Maverick's ChatGPT Secret Codes. Maverick AI, 9 Mar. 2026. PDF file. mavgpt.ai.

"Grok Imagine Prompting Guide." GenAIntel Guides, 30 Jan. 2026, www.genaintel.com.

"Grok Prompt Engineering: Full Guide." Data Studios, 21 Feb. 2026, www.datastudios.org.

Wei, Jason, et al. "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models." arXiv, 2022.

Microsoft Learn. "Prompt Engineering Techniques." Azure AI Foundry.

OpenAI. "Prompt Engineering Guide." OpenAI Platform.

Community-curated prompts and testing insights from GitHub repository "awesome-grok-imagine-prompts" and X/Reddit discussions (2026).

Share:
Expert development team

Ready to Transform Your Digital Presence?

Free consultation with our expert team

Contact Us
Contact UsorRequest a Proposal