Video Prompt: Agent Skill (SKILL.md) for Claude Code, Codex and Cursor

Writes short text-to-video or image-to-video prompts with shot, motion, timing and sound.

Prompt

Installs to .agents/skills/video-prompt-agent-skill-skill-md-for-claude-code-codex-and-curs/

mkdir -p .agents/skills/video-prompt-agent-skill-skill-md-for-claude-code-codex-and-curs && curl -sL "https://prompts-gpt.com/prompts/video-prompt-agent-skill-skill-md-for-claude-code-codex-and-cursor/raw" -o .agents/skills/video-prompt-agent-skill-skill-md-for-claude-code-codex-and-curs/SKILL.md
SKILL.md
Preview
---
name: video-prompt-agent-skill-skill-md-for-claude-code-codex-and-curs
description: "Writes short text-to-video or image-to-video prompts with shot, motion, timing and sound."
---

# Video Prompt: Agent Skill (SKILL.md) for Claude Code, Codex and Cursor

Writes short text-to-video or image-to-video prompts with shot, motion, timing and sound.

## Prompt

```
---
name: prompts-gpt-video-prompt
description: Writes short text-to-video or image-to-video prompts with shot, motion, timing and sound. Use when asked for a Veo, Sora or Kling prompt or a short video idea.
---

# Video prompt

## Steps

1. Get the brief: purpose (ad, social clip, explainer), the subject, the target model (Veo, Sora, Kling or Runway), clip length (default 8 seconds) and aspect ratio (9:16 for Shorts, Reels and TikTok; 16:9 otherwise). Ask one short question only if the subject is missing.
2. Write the shot: framing, lens and one camera move (push-in, pan, orbit or static). One continuous shot unless cuts are asked for.
3. Describe the action second by second for the clip length, with the key moment by the third second.
4. Add light, mood and sound: ambient sound, effects and any spoken line in quotes (Veo generates audio; for models that don't, put sound in a separate note).
5. For image-to-video, say what must not change from the source image and what moves.

## Output format

The prompt (under 120 words), then a timing table (second, action, sound), then one line on what to change for the other aspect ratio.

## Rules

- No real people's likeness, logos or copyrighted characters unless the user owns them.
- Physical actions only; no on-screen text longer than four words.
- State any assumption in one line at the end.
```

## Usage Notes

Save as .claude/skills/prompts-gpt-video-prompt/SKILL.md in one project (or ~/.claude/skills/prompts-gpt-video-prompt/ for all of them); Codex reads .agents/skills/prompts-gpt-video-prompt/SKILL.md. Then ask for the job in plain words and the agent loads it. Or install with: npx skills add prompts-gpt/skills --skill prompts-gpt-video-prompt.

Example output

At a glance8-second Nightjar Coffee ad, shot by shotVeo, 9:16 vertical, single continuous shot
  1. Establish0–2sBottle and iced tumbler in sunlight; push-in begins, hand lifts bottle
  2. Key moment: the pour2–4sCoffee pours over ice and swirls; ice crackle, liquid glug
  3. Cream bloom4–6sCream blooms into marbled clouds; condensation beads slide
  4. Voiceover6–7s"Slow mornings, done right."
  5. Brand hold7–8sCamera holds on glass and "Nightjar" label

Example for: a social ad for a small-batch cold brew brand "Nightjar Coffee," target model Veo, 8 seconds, 9:16 for Reels.

Prompt

Vertical 9:16, 8-second single continuous shot. Close-up at table height on a 50mm lens with shallow depth of field. A slow push-in moves toward a frosted glass bottle of dark cold brew on a weathered oak counter, lit by low amber morning sun from a side window. A hand pours the coffee over large clear ice in a tumbler. The pour slows as it swirls, and a stream of cream blooms into marbled clouds. Condensation beads roll down the glass. Warm, calm, artisanal mood with drifting dust motes. Audio: soft café ambience, the crackle of ice and a gentle glug from the pour. A relaxed female voice says, "Slow mornings, done right." The camera settles on the glass beside the bottle, which carries a small kraft label reading "Nightjar."

Timing table

SecondActionSound
0–1Static wide close-up of bottle and empty ice-filled tumbler. Sunlight is on the counter.Low café hum, distant cup clink
1–2The push-in begins. A hand lifts the bottle into frame.Faint glass-on-wood scrape
2–3Key moment: the coffee pours over the ice and the ice shifts.Ice crackle, liquid glug
3–4The pour ends and the dark coffee swirls in the glass.Gentle settling fizz
4–5Cream pours in and blooms into marbled clouds.Soft trickle
5–6The camera nears the glass. Condensation beads slide down it.Ambience swells slightly
6–7The cream swirl slows. Dust motes drift in the light.Voice: "Slow mornings, done right."
7–8The camera holds on the glass and the "Nightjar" label in soft focus.Ambience fades, final ice tick

Other aspect ratio (16:9): Widen the framing to a medium shot and place the bottle on the left third and the glass on the right third. Swap the push-in for a slow left-to-right pan across the counter.

Assumptions: "Nightjar" is the user's own brand and label. The on-screen text is limited to that single word on the bottle. The voice is a generic, unnamed narrator.

Written by AI from this prompt. Your results will vary.

Save as .claude/skills/prompts-gpt-video-prompt/SKILL.md in one project (or ~/.claude/skills/prompts-gpt-video-prompt/ for all of them); Codex reads .agents/skills/prompts-gpt-video-prompt/SKILL.md. Then ask for the job in plain words and the agent loads it. Or install with: npx skills add prompts-gpt/skills --skill prompts-gpt-video-prompt.

Make it repeatable

Keep using this prompt without pasting it in every time.

Prompt details

By
Prompts-GPT editors
Published
How prompts are written and checked

Did this prompt work?

Keep exploring

More in AI Marketing