---

name: veo-video-generator description: Generates high-fidelity 1080p videos with synced audio using Google Veo 3.1. Use for creating cinematic clips from text descriptions. metadata: clawdbot: emoji: "🎬" requires: env: ["GEMINI_API_KEY"] bins: ["node", "npm"] install: "npm install" primaryEnv: "GEMINI_API_KEY" category: "Video & Media"

Veo Video Generator

Generates short video clips with native audio using Google's state-of-the-art Veo 3.1 model.

Instructions

1. Trigger: Activate when the user wants to create or render a video.

2. Setup: The agent must run npm install once before the first execution to fetch dependencies.

3. Execution: Invoke the script by passing the user prompt as a separate argument, never by interpolating it into a shell command string. Use an argument array / execFile-style invocation so the shell never parses the prompt value. Example (Node-style pseudo-code):

   execFile('node', ['generate.js', '--prompt', userPrompt])
   

   Do not construct the command as a single concatenated string such as "node generate.js --prompt " + userPrompt.

4. Resolution: Outputs 1080p video in 9:16 aspect ratio by default.

5. Completion: Provide the user with the filename of the generated .mp4 in the workspace.

Security & Privacy

• Shell Injection Prevention: The user prompt must be passed as a discrete argument (e.g. via execFile or an argv array), never interpolated into a shell command string. Concatenating user input into a shell string (e.g. shell: true with template literals) enables shell injection and is strictly forbidden.
• Instruction Scope: This skill only sends text prompts to the Google GenAI API.
• Environment: It uses the GEMINI_API_KEY provided by the OpenClaw environment.
• Data Access: It does not read local files or .env files. All configuration is handled by the agent.