Generating Videos

Documents

generating-videos

Generating Videos is an Agent Skill for documents published in the yikart/AiToEarn repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Generates videos using Grok. Supports text-to-video and image-to-video. AI视频生成、文生视频、图生视频。

View on GitHub ↗

QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

Open your project in Codex.
Copy and paste the prompt below.
Review files and risks before approving any action.

Prompt to paste

Inspect the Agent Skill at https://github.com/yikart/AiToEarn/blob/HEAD/project/aitoearn-backend/apps/aitoearn-ai/src/core/agent/skills/generating-videos/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/generating-videos, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Video Generation

Generates videos using Grok.

Model Selection Strategy

Use grok-imagine-video for video generation.

Scenario	Use Model	Reason
Video 1-15s	`grok-imagine-video`	Grok supports direct 1-15s generation
Text-to-video	`grok-imagine-video`	Prompt-only generation
Image-to-video	`grok-imagine-video`	Uses one reference image
Video >15s	Generate multiple clips, then load `editing-videos` to concatenate	Grok max duration is 15s

Language Rule

Video prompts MUST follow the user's language. If the user writes in Chinese, generate Chinese prompts; if in English, generate English prompts. If the user explicitly requests a specific language, use that language.

Grok Video Model

Model	Speed	Quality	Max Duration	Use Case
`grok-imagine-video`	Fast	Good	15s	Default video generation

Parameters

Parameter	Values	Default	Description
`duration`	1-15	model default	Video duration in seconds
`resolution`	`480p`, `720p`	`720p`	Video resolution
`aspectRatio`	`1:1`, `16:9`, `9:16`, `4:3`, `3:4`, `3:2`, `2:3`	`9:16`	Video aspect ratio
`imageUrl`	URL	none	Reference image URL for image-to-video

Workflow

Call generateVideoWithGrok with prompt and optional parameters.
Poll getGrokVideoStatus every 30 seconds.
Maximum wait: 5 minutes.
Save completed output to draft unless the user explicitly asks for media library only.

Prompt Structure

[Subject & Background] + [Action] + [Style] + [Camera] + [Atmosphere] + [Audio]

Subject & Background

Specify the main focus and environmental context.

Examples:

"A young woman with long black hair wearing a red dress"
"White concrete apartment building with organic shapes and lush greenery"

Action

Describe what the subject is doing.

Examples:

"walks slowly towards the camera"
"transforms from liquid to solid"

Style

Add aesthetic direction.

Examples:

Film noir
Surrealism
Cyberpunk
3D cartoon animation
Cinematic documentary

Camera

Describe movement and framing.

Examples:

"slow dolly-in"
"handheld tracking shot"
"close-up with shallow depth of field"

Atmosphere

Describe lighting, color, mood, and setting.

Examples:

"warm sunset lighting"
"misty morning atmosphere"
"high contrast studio lighting"

Audio

Include useful audio cues when relevant.

Examples:

"soft ambient city sounds"
"gentle piano music"
"footsteps on wet pavement"

Aspect Ratio Guidance

Platform / Use	Aspect Ratio
TikTok, Reels, Shorts, mobile social video	`9:16`
YouTube landscape, website hero, presentation	`16:9`
Profile or square feed asset	`1:1`
Portrait-style content	`3:4` or `2:3`
Landscape social feed	`4:3` or `3:2`

Long Video Strategy

For videos longer than 15 seconds:

Split the video into independent 15-second-or-shorter segments.
Generate each segment with generateVideoWithGrok.
Poll every task until complete.
Load editing-videos.
Use the video editing tool to concatenate the completed segments.

Keep prompts visually consistent across segments by repeating subject, style, lighting, camera language, and environment details.