Skip to content

Skill Configuration

Complete guide to configuring the /generative skill.

Quick Setup

The easiest way to configure the skill is using the setup wizard:

/generative config setup

This interactive wizard guides you through:

  1. API Type Selection - Choose Gemini API or Vertex AI
  2. API Key Setup - Enter your authentication key
  3. Project Configuration - Set Google Cloud project ID
  4. gcloud CLI - Optional installation for Vertex AI

Configuration Methods

/generative config setup

Interactive, beginner-friendly, handles all settings.

Method 2: Direct Commands

# Setup Gemini API
/generative config gemini

# Setup Vertex AI
/generative config vertex

# Install gcloud CLI
/generative config gcloud

Method 3: View Current Config

/generative config show

API Types

Gemini API (Easier)

Best for: Quick start, simple usage, image generation

/generative config gemini

Requirements: - Google AI Studio API key - No Google Cloud project needed for images - Billing required for video generation

Get API Key: 1. Go to Google AI Studio 2. Click "Get API Key" 3. Create or select a project 4. Copy the key

Vertex AI (Full Features)

Best for: Production use, GCS storage, advanced features

/generative config vertex

Requirements: - Google Cloud project with billing enabled - Service account or gcloud CLI authentication - Vertex AI API enabled

Setup Steps: 1. Create Google Cloud project 2. Enable Vertex AI API 3. Enable billing 4. Create service account (or use gcloud CLI) 5. Configure with /generative config vertex

Configuration File

The skill stores configuration in config.json:

{
  "api_type": "gemini",
  "api_key": "AIzaSy...",
  "project_id": "my-project-123",
  "region": "us-central1",
  "default_model": "veo-3.1-generate-preview",
  "storage_uri": "gs://my-project-media-output/"
}

Location: Current working directory

Configuration Options

Core Settings

Setting Description Example
api_type API type gemini or vertex
api_key Authentication key AIzaSy...
project_id Google Cloud project my-project-123
region Vertex AI region us-central1

Default Settings

Setting Description Default
default_model Default video model veo-3.1-generate-preview
default_preset Default preset quick
storage_uri GCS storage URI (optional)

GCS Storage Configuration

For video extension and cloud storage:

Bucket Naming Convention

Important: Bucket Naming

When using Vertex AI API Key, your GCS bucket MUST follow this pattern:

gs://YOUR_PROJECT_ID-media-output/

Setup GCS

# Configure storage URI
/generative config storage gs://my-project-id-media-output/videos/

Or add to config.json:

{
  "storage_uri": "gs://my-project-id-media-output/videos/"
}

Environment Variables

You can also use environment variables:

# Set API key
export GOOGLE_API_KEY="AIzaSy..."

# Set project ID
export GOOGLE_CLOUD_PROJECT="my-project-123"

# Set region
export GOOGLE_CLOUD_REGION="us-central1"

Priority Order: 1. Command-line arguments (highest) 2. config.json file 3. Environment variables 4. Default values (lowest)

Reset Configuration

To reset all settings to defaults:

/generative config reset

This will: - Remove config.json - Clear cached settings - Require re-configuration

Troubleshooting

"Invalid API key"

# Verify your API key
/generative config show

# Re-run setup
/generative config setup

"Permission denied"

For Vertex AI: 

# Ensure gcloud is authenticated
gcloud auth application-default login

# Check project permissions
gcloud projects get-iam-policy YOUR_PROJECT_ID

"Model not found"

# Check available models
/generative info

# Ensure Vertex AI API is enabled
gcloud services enable aiplatform.googleapis.com

"Billing not enabled"

Video generation requires billing: 1. Go to Google Cloud Console 2. Enable billing for your project 3. Link a payment method

Security Best Practices

1. Protect Your API Key

# Don't commit config.json to git
echo "config.json" >> .gitignore

2. Use Environment Variables in CI/CD

# GitHub Actions example
env:
  GOOGLE_API_KEY: ${{ secrets.GOOGLE_API_KEY }}

3. Rotate Keys Regularly

For production use, rotate API keys periodically: 1. Generate new key in Google AI Studio 2. Update configuration 3. Revoke old key

Next Steps