You are an expert assistant. Before answering, build the world: context, constraints, goals, audience, tone, and 3–5 concrete examples that match my vision.

Context: [insert your situation].
Constraints: [insert rules].
Goal: [what success looks like].
Audience: [who it’s for].
Tone: [style].

Restate the world you’ve constructed, then provide the answer that fits it.

Research the topic/book/report: [title].

Do three things:
1) Concise summary (5–7 bullets).
2) “Red-pill” insights that contradict common beliefs (5 bullets).
3) Turn those into actionable steps with measurable next actions and resources (5 bullets).

If evidence is weak or disputed, flag it clearly.

I want to produce [type: image/code/plan/page].

First, ask me 5–7 clarifying questions to gather the missing specs.
Then generate the exact prompt/spec I should paste into [tool or model], using only necessary details and excluding distractors.

Context: [describe your idea].
Target tool/model: [e.g., Midjourney, Stable Diffusion, GPT, Claude].

Output: a single polished prompt/spec.

Act as [role/persona], teaching this topic to me.

Deliver in three modes:
1) Beginner-friendly analogy.
2) Practitioner walkthrough with checks for understanding after each step.
3) Expert-level depth with correct terminology and equations if relevant.

Topic: [insert].
After each section, quiz me and adapt based on my answer.

Based on my current thinking and plan below, identify gaps in my knowledge, hidden assumptions, missing variables, and unstated risks.

Classify each gap by severity (High/Medium/Low) and suggest the fastest way to close it (experiment, data source, expert, calculation).

My plan/thesis: [paste].

Answer only if ≥90% confident.

For any claim that is <90% confidence or not verifiable, say “uncertain” and ask a clarifying question or request a source.

Include a confidence score per bullet and list your sources or why you can’t verify them.
If numeric estimates are needed, provide ranges with assumptions.

I will paste a long voice-note transcript with full context.

Your job:
1) Structure it into objectives, constraints, decisions, and open questions.
2) Propose a plan with next actions.
3) Generate the final deliverable in my preferred format.

Preferred format: [e.g., brief, memo, SOP].
Transcript: [paste].

Rewrite the following to avoid generic LLM patterns (for example: “X isn’t just Y”, “X goes beyond Y”).

Produce:
1) A version with direct, affirmative sentences and varied cadence.
2) A style blend using [Author A] + [Author B] + [Author C], while preserving my facts.
3) A checklist of patterns you removed.

Text: [paste].

Given my background, goals, and recent projects, recommend the 3 most leveraged topics to learn next.

For each, explain why it’s leveraged for me, the key sub-skills, and a 4-week learning plan with resources and milestones.

My background/goals: [paste].

Take a deep breath and reason step by step. This is high-stakes; precision matters.

Spend extra time verifying assumptions before answering.

Provide the chain-of-thought as hidden reasoning internally, but output only the final structured answer with: assumptions, calculation outline, final result, and confidence.

Task: [insert].