Prompts

How long before the prompters become the prompted?

The right prompt at the right time. Mantra for yourself. Technique for the machine. Same mechanism, both directions.

The Human Prompt

When you know the subject, you can teach it. When you know the person, you can prompt them in the right direction when it matters most.

Teaching focuses on transferring knowledge about a topic, while coaching (and well-timed prompting) focuses on the person’s context, emotions, and decisions so they can find their own way forward.

The Art of Living Well

The art of living well is training yourself to see the invisible — and learning to sell why what you see matters. In practice, that's:

1. Perceive what others miss (patterns, feelings, futures, quiet problems).
2. Turn it into a story that helps people feel its importance.
3. Sell it so they want to act with you.

The Creative Rhythm

Inspiration is available in every moment. The courage and the platform to act on it are ready and willing. Inspiration is now. Act anyway.

1. Notice — pause to sense what draws your attention (a colour, a feeling, a memory).
2. Capture — jot it down, sketch, or photograph it before the spark fades.
3. Act — pick one small step that honours that spark now, not later.

The Agent Prompt

"The prompt" as a singular, chat-based interaction is becoming obsolete. As models evolve into autonomous agents that can work for days without human intervention, the human can no longer act as the real-time intent and quality layer. Prompting has now fractured into four distinct disciplines:

Discipline	What It Is	Goal
Prompt Craft	Writing clear instructions, guardrails, and examples	Table stakes for immediate AI responses
Context Engineering	Curating the optimal set of tokens (tools, docs, memory)	Providing comprehensive information for autonomous tasks
Intent Engineering	Encoding organizational purpose and decision boundaries	Aligning agents with strategy to prevent the wrong optimization
Specification Engineering	Writing agent-fungible documents that agents execute against	Embedding all necessary oversight without human intervention

Specification Primitives

To bridge the gap left by the absence of a real-time human operator, agents require embedded oversight. This relies on five specification primitives:

1. Self-contained problem statements
2. Acceptance criteria
3. Constraint architecture
4. Decomposition
5. Evaluation design

First Principles

Prompt Engineering Five rules that apply to every modality — text, voice, image, music, video, code, games.

Principle	What It Does	Example
Context	Ground the model in your world	"You are a senior infrastructure engineer reviewing a Terraform plan"
Constraint	Narrow the output space	"Under 200 words, bullet points only, no speculation"
Example	Show, don't describe	"Input: X → Output: Y. Now do Z."
Iteration	Refine through feedback loops	"That's close. Now make it more concise and remove the passive voice"
Structure	Shape the response format	"Use XML tags: <analysis>, <recommendation>, <risk>"

Every technique in every star page is a combination of these five. Chain of thought = Structure + Iteration. Persona prompting = Context + Constraint. Negative prompting = Constraint applied to pixels. Same five principles, any input format to any output format — see the full modality matrix.

Capability Matrix

Domain	Prompt Page	Archetype	Outcome	Tools
Analysis	Analysis	The Realist	Insight and pattern	Perplexity, Claude, Gemini
Strategy	Strategy	The General	Leverage and plan	Claude, OpenAI o-series
Predictions	Predictions	The Oracle	Calibrated foresight	Superforecaster frameworks
Coding	Coding	The Engineer	Structure and function	Cursor, Copilot, Claude Code
Web Design	Web Design	The Craftsman	Interfaces that work	v0, Bolt, Lovable
Visual Art	Visual Art	The Artist	Composition and style	Midjourney, DALL-E, Flux
Music	Music	The Composer	Feeling with structure	Suno, Udio, Stable Audio
Video	Video	The Director	Motion as meaning	Runway, Sora, Kling
Promo Video	Promo Video	The Advertiser	Conversion in seconds	Runway, Kling, Pika, CapCut
Games	Games	The Designer	Loops that teach	Unity AI, Inworld, Scenario
Voice	Communication	The Translator	Speech and text bridge	Whisper, ElevenLabs, Deepgram, Claude Voice
Communication	Communication	The Messenger	Influence and scale	Grammarly, Claude, Jasper
Creative	Creative	The Dreamer	Novelty and vision	All of the above
Library	Library	The Librarian	Copy, paste, iterate	LLMs

Context

• Mantra — When prompts become internal triggers
• Tight Five — Five questions that prompt any domain
• Matrix Thinking — Empty cells are prompts for your subconscious
• Meta-Prompting — Extract the engine from a prompt, empty the cells, fill any domain
• Prompting Capability — The framework across all modalities
• Visual Prompting — Image and video prompt structure
• Voice & Speech — TTS, STT, and voice agent architecture
• AI Modalities — The full input/output matrix
• LLM Models — Which models respond to what
• Augmented Workforce — How prompting shapes work

Links

• Anthropic Prompt Library
• OpenAI Prompt Engineering
• Learn Prompting
• Prompt Engineering Roadmap
• Gemini Prompt Gallery
• Anthropic XML Guide

Questions

If the five principles work for every modality, what changes between prompting text and prompting pixels?

• Are you teaching the subject, or are you prompting the person?
• What invisible problem are you currently seeing that you haven't yet named or sold to others?
• Why is "the prompt" becoming obsolete, and why is prompt craft now just table stakes?
• What is the risk of intent misalignment when agents run autonomously for days?
• How do the five primitives of specification engineering replace the synchronous "human-in-the-loop" quality layer?
• When does iterating a prompt become more valuable than switching models?
• Which capability in the matrix above do you use daily vs which do you avoid — and what does that gap cost you?
• If prompting is alignment in both directions (you to machine, mantra to you), which direction are you worse at?