ChatGPT Secret Codes · The AI-Powered Nonprofit

ChatGPT Secret Codes

I was in a training today and a nonprofit brought up using the ChatGPT Secret Codes.

Here is the comprehensive list of 32 shortcuts, categorized by how they help you.

Speed & Formatting (Get to the point)

Use these when you need specific output formats without the fluff.

/ELI5: Explain Like I’m 5. Great for complex concepts (Quantum physics, Blockchain).
/TLDL: "Too Long; Didn't Listen/Read". Summarizes long transcripts or texts into a few key lines.
/BRIEFLY: Forces a ruthless constraint on length. Good for quick definitions.
/EXEC SUMMARY: Generates a high-level summary suitable for a CEO or decision-maker.
/CHECKLIST: Converts the response into a functional, actionable checkbox list.
/FORMAT AS [Type]: Forces the output into a specific format (Table, JSON, Markdown, CSV).

Persona & Tone (Change the voice)

Stop the "AI voice" by forcing a specific perspective.

/ACT AS [Role]: Sets a specific persona. (e.g., /ACT AS Michelin Chef).
/TONE [Mood]: Modifies the emotional weight (Formal, Sarcastic, Urgent, Dramatic).
/AUDIENCE [Target]: Adapts complexity for a specific group (Experts, Beginners, Stakeholders).
/JARGON: Specifically asks the AI to use technical, industry-specific vocabulary (opposite of ELI5).
/DEV MODE: Simulates a raw, technical developer perspective (code-heavy, concise).
/PM MODE: Adopts a Project Manager persona (focus on timelines, resources, risks).

Deep Logic & Reasoning (Think harder)

Use these to stop hallucinations and force better logic.

/STEP-BY-STEP: Forces the AI to show its work. Proven to reduce math and logic errors.
/CHAIN OF THOUGHT: Similar to step-by-step, but focuses on the connecting logic between ideas.
/FIRST PRINCIPLES: Breaks a problem down to its fundamental truths and builds up from there.
/DELIBERATE THINKING: Forces a "slow down" approach to reasoning (great for complex strategy).
/NO AUTOPILOT: Explicit instruction to avoid generic, cliché, or lazy answers.
/REFLECTIVE MODE: Asks the AI to reflect on its own answer after generating it to check for quality.
/SYSTEMATIC BIAS CHECK: Explicitly asks the AI to scan its response for inherent biases.
/EVAL-SELF: Forces a critical self-evaluation of the response at the end.

Analysis & Strategy (Big picture thinking)

Turn the AI into a business consultant.

/SWOT: Generates a Strengths, Weaknesses, Opportunities, and Threats analysis.
/COMPARE: Puts two or more concepts side-by-side (best combined with /FORMAT AS TABLE).
/MULTI-PERSPECTIVE: Explores a topic from 3-4 different viewpoints (e.g., Economic, Social, Ethical).
/PARALLEL LENSES: Similar to multi-perspective, but examines a singular issue through specific theoretical lenses.
/PITFALLS: Specifically focuses on what could go wrong or common mistakes in a plan.
/METRICS MODE: Forces the answer to include measurable KPIs, numbers, or success indicators.
/CONTEXT STACK: Instructs the AI to keep previous context layers active (useful for long chats).

Advanced Control (The power user tools)

/ROLE: TASK: FORMAT:: The "God Mode" of prompts. Defines everything in one line.

/SCHEMA: Generates a structured outline or data model for a project.
/REWRITE AS [Style]: Takes existing text and transforms it (e.g., /REWRITE AS Seinfeld script).
/BEGIN WITH / END WITH: Constraints the AI to start or end sentences in a specific way (great for coding or creative writing constraints).
/GUARDRAIL: Sets strict negative constraints (e.g., "Do not use emojis", "Do not mention X").

Here is a list of them to reference later.

0 comments

ChatGPT Secret Codes

skool.com/ai-clarity-for-nonprofits-1475

Build an AI-powered funding & donor retention engine inside Google Workspace. Stop donor churn & automate grants without expensive software

Suggested communities

Build your own community

Bring people together around your passion and get paid.