veo3-prompter_skill

This skill crafts professional Veo 3.1 prompts with cinematic structure, audio direction, and timestamp choreography to produce polished AI videos.
  • Python

20

GitHub Stars

4

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill leegonzales/aiskills --skill veo3-prompter

  • CHANGELOG.md1.4 KB
  • LICENSE1.0 KB
  • README.md4.0 KB
  • SKILL.md8.0 KB

Overview

This skill crafts professional prompts for Google Veo 3.1 that combine cinematic structure, audio direction, and timestamp choreography. It translates a creative idea into a compact, high-control prompt format that Veo can execute reliably. The output emphasizes shot composition, prioritized detail, and synchronized sound for short AI-generated video clips.

How this skill works

I apply a five-part formula—Cinematography + Subject + Action + Context + Style & Audio—to produce prompts that balance guidance and freedom. Prompts are calibrated using a three-tier priority system so essential elements are always present while avoiding overload. For multi-shot sequences I produce timestamped choreography (up to 8 seconds) with explicit SFX, dialogue, and ambient layers. I also format quick one-liners, structured multi-line prompts, and storyboard-style sets for iterative work.

When to use it

  • Generating AI videos with Veo 3 or Veo 3.1
  • Creating multi-shot, cinematic or documentary-style short clips
  • Preparing timestamped choreography for synchronized sound and action
  • Prototyping scenes that need repeatable output with a fixed seed
  • Translating a script or concept into production-ready prompt instructions

Best practices

  • Always include Tier 1 elements: shot size, subject identity, primary action, and a dominant mood/style word
  • Add one camera movement OR one camera angle, plus one lighting note from Tier 2
  • Start with Tier 1 only for tests, then add Tier 2 items that matter most
  • Limit Tier 3 details to 1–2 items to avoid over-constraining the model
  • Use explicit audio labels: Dialogue in quotes, SFX:, Ambient:, Music:
  • When iterating, keep seed constant to reproduce results and change one variable at a time

Example use cases

  • Quick Prompt: one-line cinematic shot for a marketing teaser
  • Structured Prompt: multi-line scene with specified camera move and single SFX
  • Timestamp Sequence: 4–8s choreographed cut with per-shot audio cues
  • Storyboard Mode: multiple related prompts for sequential scene planning
  • Nature doc-style close-up with macro lens, ambient birdsong, and slow motion

FAQ

Aim for concise prompts that include all Tier 1 elements and 1–2 Tier 2 items; single-line quick prompts or structured 3–6 line prompts work best.

What if the model ignores instructions?

Remove the least important detail, rerun with Tier 1+critical Tier 2, or use a seed and iterate changing one element at a time.

Can I generate longer scenes?

Veo 3.1 supports short outputs (4, 6, or 8 seconds); for longer narratives use storyboard mode or stitch multiple generations together.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
veo3-prompter skill by leegonzales/aiskills | VeilStrat