Skip to content

CONTRIBUTING.md

Latest commit

聽

History

History
46 lines (29 loc) 路 2.21 KB

CONTRIBUTING.md

File metadata and controls

46 lines (29 loc) 路 2.21 KB
Raw

Contributing to Awesome Gemini Omni 1.1 Flash Prompts

Contributions should make a real video task easier to direct, reproduce, and review. Submit original work that you have the right to share.

Prompt entry format

Use a stable category ID and include every field needed to execute the idea:

<a id="cat-01"></a>

## CAT-01 | Clear, searchable title

**Workflow:** Text to video 路 **Inputs:** none

```text
Complete copy-ready prompt with scene, ordered action, camera, sound, continuity constraints, and ending state.
```

**Adjust:** useful variables. **Review:** observable acceptance criteria.

Quality checklist

  • Describe one coherent goal instead of stacking unrelated visual keywords.
  • Assign one role to every image or video reference.
  • State what changes and what remains unchanged in editing prompts.
  • Keep identity anchors, object geometry, screen direction, and input order consistent.
  • Identify the speaker, exact dialogue, ambience, Foley, music, or intentional silence.
  • Make the ending state observable and hold it long enough to inspect when relevant.
  • Use scene-specific exclusions rather than a generic negative-word list.
  • Separate creative intent from interface fields and documented model limits.

Rights and safety

Do not submit API keys, private documents, signed media URLs, unauthorized likenesses, protected characters, unlicensed media, tracking links, or promotional copy. Avoid imitation of a living artist's style. When factual, medical, scientific, historical, or safety-sensitive material appears in a prompt, require qualified review of the final output.

Evidence

If you include a result, identify the actual model, surface, settings, date, input roles, and meaningful observations. A polished frame does not demonstrate temporal consistency, audio accuracy, or successful editing across the complete clip.

Media submissions

For every new image or video, include its public path, source or generation tool, complete prompt, generation date, meaningful edits, input rights, and linked prompt ID. Do not present concept artwork as a target-model result.

By contributing, you confirm that you can share the submitted text and media under this repository's MIT License.