skills.rest
Aadi-110i
Aadi-110iCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

video-prompting-guide

Provides prompt structure formulas and keyword references for AI video generation models.

Updated Jul 28, 2026
Search Tags:#prompt engineering#video generation#cinematography#ai video#seedance#prompt guide#veo
One-click install
npx skills add https://github.com/Aadi-110i/PEP-PROJECT --skill video-prompting-guide-aadi-110i
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: video-prompting-guide
Source: https://github.com/Aadi-110i/PEP-PROJECT/tree/main/skills/video-prompting-guide
Command: npx skills add https://github.com/Aadi-110i/PEP-PROJECT --skill video-prompting-guide-aadi-110i

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writing effective prompts for AI video generators like Veo, Seedance, Wan, Grok, Kling, Runway, Pika, and Sora is difficult because vague or poorly structured prompts produce inconsistent, low-quality results. This Skill provides a proven prompt structure formula, keyword libraries, and model-specific tips to get consistent, professional video output. ## Core Features & Use Cases - Prompt Structure Formula: A repeatable template combining shot type, subject, action, setting, lighting, style, and technical keywords for predictable results. - Keyword Reference Tables: Curated lists of shot types, camera movements, lighting terms, and style keywords mapped to their visual effects. - Model-Specific Guidance: Tailored tips for Google Veo, Seedance, Wan 2.5, and Grok, plus ready-to-run CLI examples via the inference.sh belt CLI. - Use Case: A content creator needs a product demo video. Instead of guessing, they apply the formula—"Smooth tracking shot around a sleek smartphone on a white pedestal, soft studio lighting, 4K, shallow depth of field"—and run it through Veo for a polished result. ## Quick Start Ask the AI to write a cinematic video prompt for your subject using the shot type, lighting, and style formula, then run it with the belt CLI against a model like google/veo-3-1-fast.

Frequently Asked Questions about video-prompting-guide

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a good AI video generation prompt?▼

Structure your prompt as shot type plus subject, action, setting, lighting, style, and technical keywords. For example: "Slow motion close-up of coffee being poured, morning sunlight, warm color grading, cinematic, 4K, shallow depth of field."

What is the best prompt structure for Veo and Seedance?▼

Use the formula: shot type, subject, action, setting, lighting, style, then technical quality terms. Veo excels at realistic cinematic content with detailed prompts, while Seedance handles human motion and dance well and supports first-frame control.

Which camera movement keywords work in AI video prompts?▼

Common effective keywords include tracking shot, dolly in/out, pan, tilt, crane shot, handheld, steadicam, zoom, and static. Each produces a distinct effect, such as tracking shots for dynamic following or static shots for contemplative scenes.

Why do my AI video prompts give inconsistent results?▼

Inconsistency usually comes from vague descriptions, multiple scenes in one prompt, missing motion cues, or conflicting style keywords. Fix this by specifying one scene, adding camera movement or action, and choosing a single coherent aesthetic.

Can I use image-to-video with these prompting techniques?▼

Yes, Wan 2.5 is designed for image-to-video and animates still images naturally with good motion prediction. Seedance also supports first-frame control and works well with reference images for consistent character motion.

What are common mistakes in AI video prompting?▼

The main mistakes are being too vague, cramming multiple scenes into one prompt, omitting motion, mixing conflicting styles like minimalist and maximalist, and leaving lighting undefined. Each mistake has a direct fix in the prompt formula.

Related Skills

Explore More Design & Creative Skills→
contentbugvideoeditingcontentbugvideoediting

fal-ai-media

Generate images, videos, and audio from prompts via fal.ai MCP.

Community
Advanced
inference-shinference-sh
723

seedance

Generate videos with synchronized audio using ByteDance Seedance 2.0 via the inference.sh CLI.

Official
Intermediate
psw7205psw7205
1

prompt

Diagnoses and refines prompts by improving structure, techniques, and wording.

Community
Intermediate
SheshiyerSheshiyer
4

image-to-video

Convert still images into animated videos using the infsh CLI and motion models.

Community
Intermediate