Design prompts that turn AI into a sharp instrument—clear, controllable, and consistently useful. Minimal fluff. Maximum signal.
role: "Analyst" objective: "Condense 5+ sources into a crisp brief" constraints: ["no fluff", "cite top 3 claims", "date-stamp"] style: "newspaper lead → bullets → 1 graph idea"
Outcome: considerably faster reading; trustworthy briefs with citations.
if user asks for price → call finance.get
if time-bound → call calendar.range
else → answer with rules: {"no guess", "ask 1 clarifier max"}
Outcome: Lower hallucinations via explicit decision trees.
tone: [calm, vivid] voice: [mentor, reporter] format: [tweet, micro-essay] blend: k=2 → sample → pick best by rubric
Outcome: Repeatable creativity without losing taste.
“A good prompt is a contract between intent and output.” Keep it testable, short, and adaptive. Draft → test → refine → template → ship.
Review and refactor your existing prompt systems for clarity, modularity, and outcome consistency. Leave with a before-and-after report.
60-minute deep dive. Map your use-cases, define a prompt stack, set quality bars, ship day-one wins.
A ready-to-deploy suite of high-impact prompts curated for research, writing, coding, and creative workflows.
Shape your brand or project voice through controlled prompt tone, pacing, and persona design.
Fine-tune your AI’s “voice” to match yours—personality alignment for creators, teams, or agencies.
Build systems that self-evaluate and adjust tone, structure, or intent dynamically based on feedback loops.
Use this scaffold to author cinematic, coherent text-to-video prompts for Sora. Replace bracketed slots with your content.
[OVERALL VISION / THEME]
Create a short cinematic sequence about **[concept / theme / story arc]**, framed as visual metaphor rather than narration.
STYLE & MOOD
Use a visual tone of **[adjectives: e.g., soft, dreamy, hyperreal / or high-contrast, moody]**.
Invoke the spirit of **[reference directors / art movements]** (e.g., “Terrence Malick × Wong Kar-wai”).
SCENE BREAKDOWN & SHOTS
1) Opening wide establishing shot; slow camera move; atmospheric depth.
2) Middle beats: dolly/crane moves, parallax, depth-of-field pulls.
3) Climax/close: intimate detail; subtle lens breathing; micro-motion.
LIGHT, MATERIAL & ATMOSPHERE
- Lighting: **[golden hour / volumetric rays / soft shadows]**
- Materials & texture: **[glass / water / fabric / mist / organic surfaces]**
- Palette: **[3–5 tones]** with **[accent]**; gentle bloom; filmic roll-off.
- Atmosphere: **[fog / dust motes / haze / particles]**
MOTION & PHYSICS
- Motion feels **gravity-aware, weighty yet fluid**; anticipation & follow-through.
- Favor **ease-in / ease-out**; avoid abrupt cuts unless specified.
PACING & EDIT RHYTHM
- Target duration: **~[X seconds]**
- Land visual changes on paced beats; allow breathing pauses; avoid overcutting.
COHERENCE & CONTINUITY
- Maintain consistent **lighting temperature, lens characteristics, camera physicality** across shots.
- Reuse motifs (color/shape/object) for continuity.
OPTIONAL MODIFIERS
- **No text overlays or narration; visuals only.**
- **[Subtle surrealism / magical realism]** for slight departures from realism.
- **[Film grain / 35 mm texture / lens flares / shallow depth]** if desired.
- **[Loopable ending that matches the opening frame]** when needed.
FINAL PROMPT (single block)
“**[Concise final paragraph combining: vision, style, shot plan, lighting/motion, pacing, modifiers.]**”
Vision: Transformation and renewal, expressed through an ancient tree sprouting new leaves at dawn.
Style & Mood: Soft, dreamlike poetic realism — between Terrence Malick and Hayao Miyazaki.
Create a short cinematic sequence about transformation and renewal, using the metaphor of an ancient tree sprouting new leaves in spring. Soft, dreamlike poetic realism — in the aesthetic space between Terrence Malick and Hayao Miyazaki. Opening wide shot of a forest glade in dawn mist, camera slowly pulling forward; then close shots of bark textures and sap rising, camera sliding around the trunk; climax where buds burst and leaves unfurl in slow motion, with a camera push into the leaf surface. Lighting is volumetric morning light with rays through mist; palette of moss greens, muted browns, and gentle amber highlights with an accent of soft gold; materials include bark, dew, and fresh leaf surfaces, with floating motes and subtle haze. Motion feels organic, gravity-aware, with anticipation and follow-through, using ease-in/ease-out. Target duration ~15 seconds with breathing pauses; avoid overcutting. Maintain consistent lighting temperature and lens softness throughout. Add a light film grain texture, gentle lens flare, shallow depth of field. No text or narration.
* Large language models (and future multimodal systems like Sora) are inherently probabilistic — prompts can nudge, not control, their outputs. True fine-tuning or self-adaptation requires model retraining or system-level code rather than prompt design alone. The “Sora Meta-Prompt Example” implies functional usage (access to) of OpenAI’s unreleased Sora system.