← Back to Skills Marketplace
33
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install vertex-video-adc
Description
Generate videos via Google Cloud Vertex AI predictLongRunning (Veo) using Application Default Credentials (ADC).
README (SKILL.md)
Google Cloud Vertex AI Video Generator (ADC)
This skill utilizes the Google Cloud SDK (gcloud) and application credentials to call the Google Cloud Vertex AI predictLongRunning (Veo) endpoint, generating videos from text prompts or reference images.
Workflow
- Configure required environment variables or parameters:
GOOGLE_CLOUD_PROJECT: Your Google Cloud Project ID.GOOGLE_CLOUD_LOCATION(Optional): Location/region (defaults tous-central1).GOOGLE_CLOUD_MODEL(Optional): Model name (defaults toveo-3.1-generate-001).
- Construct/save prompt to a local text file.
- Call the generation script.
Commands
Text-to-Video
python3 {baseDir}/scripts/video_harness.py --prompt-file \x3Cprompt_path> --out \x3Coutput_path>
Image-to-Video
Provide an optional reference image:
python3 {baseDir}/scripts/video_harness.py --prompt-file \x3Cprompt_path> --image \x3Cimage_path> --out \x3Coutput_path>
Advanced Usage
- Specify custom project, location, and model:
python3 {baseDir}/scripts/video_harness.py --prompt-file \x3Cprompt_path> --out \x3Coutput_path> --project \x3Cproject_id> --location \x3Clocation> --model \x3Cmodel>
- Custom resolution/aspect ratio/duration/audio generation:
python3 {baseDir}/scripts/video_harness.py --prompt-file \x3Cprompt_path> --out \x3Coutput_path> --aspect-ratio 16:9 --duration 5 --resolution 720p --generate-audio false
Setup & Requirements
- Google Cloud SDK (
gcloudCLI) installed and available in$PATH. - Active authentication token:
- Default ADC mode:
gcloud auth application-default login - User-token mode:
gcloud auth login
- Default ADC mode:
- The system must have
curlinstalled to issue the API request.
Usage Guidance
Review this before installing on shared machines or when prompts, source images, or generated videos are sensitive. The Google Cloud/API behavior is expected, but the script should ideally remove default debug persistence or write any debug files to private, per-run locations with restrictive permissions.
Capability Assessment
Purpose & Capability
The stated purpose is Vertex AI video generation, and the script's use of prompts, optional reference images, Google Cloud authentication, curl, and output video files is coherent with that purpose.
Instruction Scope
The user-facing instructions disclose gcloud, curl, ADC/user-token authentication, prompt files, optional image input, and output paths, but they do not mention that raw API response data is also written locally for debugging.
Install Mechanism
The artifact contains README.md, SKILL.md, and one Python script; there is no installer, startup hook, background service, or bundled obfuscated code.
Credentials
Google Cloud credentials and project/location environment variables are proportionate for Vertex AI generation, but users should understand that the skill obtains an access token through gcloud and sends prompt/image content to Google Cloud.
Persistence & Privilege
The script writes predictLongRunning, polling, and final operation responses to fixed filenames in the system temp directory by default, without a debug flag, cleanup, randomized names, or disclosure in the skill instructions.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install vertex-video-adc - After installation, invoke the skill by name or use
/vertex-video-adc - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release: Generate videos using Google Cloud Vertex AI (Veo) via Application Default Credentials.
- Supports text-to-video and image-to-video generation from prompt files.
- Uses the `gcloud` CLI and `curl` for API access and authentication.
- Allows custom project, location, model, resolution, aspect ratio, duration, and audio settings.
- Requires Google Cloud SDK and `curl` to be installed.
- Simple CLI workflow for prompt handling and video generation.
Metadata
Frequently Asked Questions
What is Google vertex video adc?
Generate videos via Google Cloud Vertex AI predictLongRunning (Veo) using Application Default Credentials (ADC). It is an AI Agent Skill for Claude Code / OpenClaw, with 33 downloads so far.
How do I install Google vertex video adc?
Run "/install vertex-video-adc" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Google vertex video adc free?
Yes, Google vertex video adc is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Google vertex video adc support?
Google vertex video adc is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Google vertex video adc?
It is built and maintained by 0YU (@neogeosnk); the current version is v1.0.0.
More Skills