Initial commit

2025-11-30 08:21:41 +08:00
commit 0b3ddcd76a
8 changed files with 603 additions and 0 deletions
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -0,0 +1,15 @@
+{
+  "name": "003-jeremy-vertex-ai-media-master",
+  "description": "Comprehensive Google Vertex AI multimodal mastery for Jeremy - video processing (6+ hours), audio generation, image creation with Gemini 2.0/2.5 and Imagen 4. Marketing campaign automation, content generation, and media asset production.",
+  "version": "1.0.0",
+  "author": {
+    "name": "Jeremy Longshore",
+    "email": "jeremy@intentsolutions.io"
+  },
+  "skills": [
+    "./skills"
+  ],
+  "commands": [
+    "./commands"
+  ]
+}
--- a/README.md
+++ b/README.md
@@ -0,0 +1,3 @@
+# 003-jeremy-vertex-ai-media-master
+
+Comprehensive Google Vertex AI multimodal mastery for Jeremy - video processing (6+ hours), audio generation, image creation with Gemini 2.0/2.5 and Imagen 4. Marketing campaign automation, content generation, and media asset production.
--- a/commands/vertex-campaign.md
+++ b/commands/vertex-campaign.md
@@ -0,0 +1,167 @@
+---
+name: vertex-campaign
+description: Generate complete multimodal marketing campaigns using Vertex AI - video, audio, images, copy for all channels
+model: sonnet
+---
+
+# Generate Multimodal Marketing Campaign with Vertex AI
+
+Create a comprehensive marketing campaign with all assets generated via Google Vertex AI multimodal capabilities.
+
+## What This Does
+
+1. **Campaign Brief Analysis**: Understand product, target audience, goals
+2. **Asset Generation**: Create all required media assets
+3. **Multi-Channel Content**: Generate content for all marketing channels
+4. **Implementation Guide**: Provide deployment instructions
+
+## Campaign Assets Generated
+
+### Visual Assets (Imagen 4)
+- Hero image (1920x1080)
+- Social media graphics (Instagram, Facebook, LinkedIn)
+- Display ad creatives (multiple sizes)
+- Product lifestyle images
+- A/B test variations
+
+### Video Assets (Gemini 2.5 Pro)
+- Video scripts (30s, 60s, 2min versions)
+- Storyboard descriptions
+- Video editing instructions
+- Thumbnail designs
+
+### Audio Assets (Lyria)
+- Background music compositions
+- Voiceover scripts
+- Audio ad scripts
+- Podcast episode outlines
+
+### Written Content (Gemini 2.5 Pro)
+- Email marketing sequences
+- Blog post (SEO optimized)
+- Social media captions
+- PPC ad copy
+- Landing page copy
+
+## Usage
+
+```bash
+/vertex-campaign
+```
+
+Then provide campaign details:
+- Product/service name
+- Target audience
+- Campaign objectives
+- Brand guidelines
+- Budget considerations
+
+## Example Workflow
+
+**Input:**
+```
+Product: Premium noise-canceling headphones
+Audience: Remote workers, 25-45, tech-savvy
+Goal: Product launch, 10K units in Q1
+Budget: $50K
+```
+
+**Output:**
+1. 15 product images (lifestyle, studio, use-cases)
+2. 30s product launch video script
+3. Background music track (energetic, professional)
+4. Email sequence (5 emails)
+5. Social media content (30 posts across platforms)
+6. Blog post "Best Headphones for Remote Work 2025"
+7. PPC campaigns (Google, Facebook, LinkedIn)
+
+## Technical Implementation
+
+**Step 1: Initialize Vertex AI**
+```python
+from google.cloud import aiplatform
+from vertexai.preview.generative_models import GenerativeModel
+from vertexai.preview.vision_models import ImageGenerationModel
+
+aiplatform.init(project=PROJECT_ID, location="us-central1")
+```
+
+**Step 2: Generate Visual Assets**
+```python
+imagen = ImageGenerationModel.from_pretrained("imagen-4")
+hero_image = imagen.generate_images(
+    prompt=f"Professional product photography of {product}, studio lighting, clean background",
+    number_of_images=1,
+    aspect_ratio="16:9"
+)
+```
+
+**Step 3: Create Video Script**
+```python
+gemini = GenerativeModel("gemini-2.5-pro")
+video_script = gemini.generate_content([
+    f"Create a 30-second video script for {product} targeting {audience}. Include scene descriptions, voiceover, music cues."
+])
+```
+
+**Step 4: Generate Audio**
+```python
+from vertexai.preview.audio_models import AudioGenerationModel
+lyria = AudioGenerationModel.from_pretrained("lyria")
+background_music = lyria.generate_audio(
+    prompt=f"Background music for {product} video ad, {mood}, 30 seconds",
+    duration=30
+)
+```
+
+**Step 5: Create Multi-Channel Copy**
+```python
+content = gemini.generate_content([
+    f"""Generate marketing content for {product}:
+    - 5-email drip campaign
+    - 10 Instagram captions
+    - 5 LinkedIn posts
+    - SEO blog post (1500 words)
+    - Google Ads copy (5 variations)"""
+])
+```
+
+## Cost Estimation
+
+**Per Campaign:**
+- Visual Assets: $2-3 (50 images @ $0.04 each)
+- Video Scripts: $0.50 (Gemini tokens)
+- Audio: $1-2 (Lyria generation)
+- Written Content: $1 (Gemini tokens)
+
+**Total: ~$5-7 per complete campaign**
+
+## Best Practices
+
+1. **Brand Consistency**: Provide brand guidelines in prompt
+2. **Batch Generation**: Generate multiple variations simultaneously
+3. **Quality Control**: Review and iterate on generated assets
+4. **Version Control**: Save prompts and outputs for reproducibility
+5. **A/B Testing**: Generate 3-5 variations of each asset
+
+## Integration with Marketing Stack
+
+**Export to:**
+- Google Ads (PMax campaigns)
+- Meta Business Suite (Facebook/Instagram)
+- LinkedIn Campaign Manager
+- Email marketing platforms (HubSpot, Mailchimp)
+- CMS platforms (WordPress, Contentful)
+
+## Performance Tracking
+
+**Monitor:**
+- Asset generation time
+- Cost per asset
+- Approval rates
+- Campaign performance metrics
+- ROI vs traditional production
+
+---
+
+**This command turns Jeremy into a one-person marketing agency powered by Vertex AI multimodal capabilities.**
--- a/plugin.lock.json
+++ b/plugin.lock.json
@@ -0,0 +1,61 @@
+{
+  "$schema": "internal://schemas/plugin.lock.v1.json",
+  "pluginId": "gh:jeremylongshore/claude-code-plugins-plus:plugins/productivity/003-jeremy-vertex-ai-media-master",
+  "normalized": {
+    "repo": null,
+    "ref": "refs/tags/v20251128.0",
+    "commit": "bafc37dab547b5448ac87ca6c3ce12b1dfaad7f5",
+    "treeHash": "5a4b2df6319d79d9e8c9738ef75cd0e41376c858af9d997ce38b2ed47da67379",
+    "generatedAt": "2025-11-28T10:18:02.066416Z",
+    "toolVersion": "publish_plugins.py@0.2.0"
+  },
+  "origin": {
+    "remote": "git@github.com:zhongweili/42plugin-data.git",
+    "branch": "master",
+    "commit": "aa1497ed0949fd50e99e70d6324a29c5b34f9390",
+    "repoRoot": "/Users/zhongweili/projects/openmind/42plugin-data"
+  },
+  "manifest": {
+    "name": "003-jeremy-vertex-ai-media-master",
+    "description": "Comprehensive Google Vertex AI multimodal mastery for Jeremy - video processing (6+ hours), audio generation, image creation with Gemini 2.0/2.5 and Imagen 4. Marketing campaign automation, content generation, and media asset production.",
+    "version": "1.0.0"
+  },
+  "content": {
+    "files": [
+      {
+        "path": "README.md",
+        "sha256": "d275f21f45d75faf2e8248e842a9e1bd1a9290c8985c46512f2348725ed55753"
+      },
+      {
+        "path": ".claude-plugin/plugin.json",
+        "sha256": "eb6375d110ebb4662f0d65f3d7f939d925354e7eb42f075a86c989a8e6c32a0f"
+      },
+      {
+        "path": "commands/vertex-campaign.md",
+        "sha256": "e7f0161ca2bffa3cacd4caa706aed28a941587b3a758ce486d3393c3cea47d5a"
+      },
+      {
+        "path": "skills/vertex-media-master/SKILL.md",
+        "sha256": "b8562b0758320a9bb48d42f571672d0cdc79bac1bea369a2f116a085922a5a73"
+      },
+      {
+        "path": "skills/vertex-media-master/references/README.md",
+        "sha256": "db9680278e03728fef93321fc76c435387bc0c8fe1dcc9870bdf2fa236ea8ac3"
+      },
+      {
+        "path": "skills/vertex-media-master/scripts/README.md",
+        "sha256": "f042646ad5b685556c044080a6b73202a490fb8288be8219328faefc12d5a30e"
+      },
+      {
+        "path": "skills/vertex-media-master/assets/README.md",
+        "sha256": "33bfb083485b48c78a1738368c52cd9f202724a414bce507db181d8291b83aec"
+      }
+    ],
+    "dirSha256": "5a4b2df6319d79d9e8c9738ef75cd0e41376c858af9d997ce38b2ed47da67379"
+  },
+  "security": {
+    "scannedAt": null,
+    "scannerVersion": null,
+    "flags": []
+  }
+}
--- a/skills/vertex-media-master/SKILL.md
+++ b/skills/vertex-media-master/SKILL.md
@@ -0,0 +1,281 @@
+---
+name: Vertex AI Media Master
+description: |
+  Automatic activation for ALL Google Vertex AI multimodal operations - video processing, audio generation, image creation, and marketing campaigns.
+  **TRIGGER PHRASES:**
+  - "vertex ai", "gemini multimodal", "process video", "generate audio", "create images", "marketing campaign"
+  - "imagen", "video understanding", "multimodal", "content generation", "media assets"
+  **AUTO-INVOKES FOR:**
+  - Video processing and understanding (up to 6 hours)
+  - Audio generation and transcription
+  - Image generation with Imagen 4
+  - Marketing campaign automation
+  - Social media content creation
+  - Ad creative generation
+  - Multimodal content workflows
+allowed-tools: Read, Write, Edit, Grep, Glob, Bash
+version: 1.0.0
+---
+
+# Vertex AI Media Master - Comprehensive Multimodal AI Operations
+
+This Agent Skill provides comprehensive mastery of Google Vertex AI multimodal capabilities for video, audio, image, and text processing with focus on marketing applications.
+
+## Core Capabilities
+
+### 🎥 Video Processing (Gemini 2.0/2.5)
+- **Video Understanding**: Process videos up to 6 hours at low resolution or 2 hours at default resolution
+- **2M Context Window**: Gemini 2.5 Pro handles massive video content
+- **Audio Track Processing**: Automatic audio transcription from video
+- **Multi-video Analysis**: Process multiple videos in single request
+- **Video Summarization**: Extract key moments, scenes, and insights
+- **Marketing Use Cases**:
+  - Analyze competitor video ads
+  - Extract highlights from long-form content
+  - Generate video summaries for social media
+  - Transcribe and caption video content
+  - Identify brand mentions and product placements
+
+### 🎵 Audio Generation & Processing
+- **Lyria Model (2025)**: Native audio and music generation
+- **Speech-to-Text**: Transcribe audio with speaker diarization
+- **Text-to-Speech**: Generate natural voiceovers
+- **Music Composition**: Background music for campaigns
+- **Audio Enhancement**: Noise reduction and quality improvement
+- **Marketing Use Cases**:
+  - Generate podcast scripts and voiceovers
+  - Create audio ads and radio spots
+  - Produce background music for video campaigns
+  - Transcribe customer interviews
+  - Generate multilingual voiceovers
+
+### 🖼️ Image Generation (Imagen 4 & Gemini 2.5 Flash Image)
+- **Imagen 4**: Highest quality text-to-image generation
+- **Gemini 2.5 Flash Image**: Interleaved image generation with text
+- **Style Transfer**: Apply brand styles to generated images
+- **Product Visualization**: Generate product mockups
+- **Campaign Assets**: Create ad creatives and social media graphics
+- **Marketing Use Cases**:
+  - Generate personalized ad images (Adios solution)
+  - Create social media graphics at scale
+  - Produce product lifestyle images
+  - Generate A/B test variations
+  - Create branded campaign visuals
+
+### 📢 Marketing Campaign Automation
+- **ViGenAiR**: Convert long-form video ads to short formats automatically
+- **Adios**: Generate personalized ad images tailored to audience context
+- **Campaign Asset Generation**: Photos, soundtracks, voiceovers from prompts
+- **Content Pipeline**: Email copy, blog posts, social media, PMax assets
+- **Catalog Enrichment**: Multi-agent workflow for product onboarding
+- **Marketing Use Cases**:
+  - Automated campaign asset production
+  - Personalized content at scale
+  - Multi-channel content distribution
+  - Product catalog enhancement
+  - Visual merchandising automation
+
+### 🔧 Technical Implementation
+
+**API Integration:**
+```python
+from google.cloud import aiplatform
+from vertexai.preview.generative_models import GenerativeModel
+
+# Initialize Vertex AI
+aiplatform.init(project="your-project", location="us-central1")
+
+# Gemini 2.5 Pro for video
+model = GenerativeModel("gemini-2.5-pro")
+
+# Process video with audio
+response = model.generate_content([
+    "Analyze this video and extract key marketing insights",
+    video_file,  # Up to 6 hours
+])
+
+# Imagen 4 for image generation
+from vertexai.preview.vision_models import ImageGenerationModel
+imagen = ImageGenerationModel.from_pretrained("imagen-4")
+images = imagen.generate_images(
+    prompt="Professional product photo, studio lighting, white background",
+    number_of_images=4
+)
+```
+
+**Gemini 2.5 Flash Image (Interleaved Generation):**
+```python
+# Generate images within text responses
+model = GenerativeModel("gemini-2.5-flash-image")
+response = model.generate_content([
+    "Create a 5-step recipe with images for each step"
+])
+# Returns text + images interleaved
+```
+
+**Audio Generation (Lyria):**
+```python
+from vertexai.preview.audio_models import AudioGenerationModel
+lyria = AudioGenerationModel.from_pretrained("lyria")
+audio = lyria.generate_audio(
+    prompt="Upbeat background music for product launch video, 30 seconds",
+    duration=30
+)
+```
+
+### 📊 Marketing Workflow Automation
+
+**1. Multi-Channel Campaign Creation:**
+```python
+# Single prompt generates all assets
+campaign = model.generate_content([
+    """Create a product launch campaign for [product]:
+    - Hero image (1920x1080)
+    - 3 social media graphics (1080x1080)
+    - 30-second video script
+    - Background music description
+    - Email marketing copy
+    - Instagram caption"""
+])
+```
+
+**2. Video Repurposing Pipeline:**
+```python
+# Long-form to short-form conversion (ViGenAiR approach)
+long_video = "gs://bucket/original-ad-60s.mp4"
+response = model.generate_content([
+    f"Extract 3 engaging 15-second clips from this video for TikTok/Reels",
+    long_video
+])
+# Auto-generates format-specific versions
+```
+
+**3. Personalized Ad Generation:**
+```python
+# Context-aware image generation (Adios approach)
+for audience in audiences:
+    ad_image = imagen.generate_images(
+        prompt=f"Product ad for {product}, targeting {audience.demographics}, {audience.style_preference}",
+        aspect_ratio="16:9"
+    )
+```
+
+### 🎯 Best Practices for Jeremy
+
+**1. Project Setup:**
+```bash
+# Set environment variables
+export GOOGLE_CLOUD_PROJECT="your-project-id"
+export GOOGLE_APPLICATION_CREDENTIALS="path/to/service-account.json"
+
+# Install SDK
+pip install google-cloud-aiplatform[vision,audio] google-generativeai
+```
+
+**2. Rate Limits & Quotas:**
+- Gemini 2.5 Pro: 2M tokens/min (video processing)
+- Imagen 4: 100 images/min
+- Monitor usage in Cloud Console
+
+**3. Cost Optimization:**
+- Use Gemini 2.5 Flash for faster, cheaper operations
+- Batch image generation requests
+- Cache video embeddings for repeated analysis
+- Use low-resolution video setting when appropriate
+
+**4. Security & Compliance:**
+- Keep API keys in Secret Manager, never in code
+- Use service accounts with minimal permissions
+- Enable VPC Service Controls for data residency
+- Log all API calls for audit trails
+
+### 🚀 Advanced Marketing Use Cases
+
+**1. Campaign Performance Analysis:**
+```python
+# Analyze competitor campaigns
+competitor_videos = ["gs://bucket/competitor1.mp4", "gs://bucket/competitor2.mp4"]
+analysis = model.generate_content([
+    "Compare these competitor videos: themes, messaging, CTAs, production quality",
+    *competitor_videos
+])
+```
+
+**2. Content Localization:**
+```python
+# Generate multilingual campaigns
+for lang in ["en", "es", "fr", "de", "ja"]:
+    localized_content = model.generate_content([
+        f"Translate and culturally adapt this campaign for {lang} market:",
+        campaign_brief,
+        hero_image
+    ])
+```
+
+**3. A/B Test Generation:**
+```python
+# Generate variations automatically
+variations = []
+for style in ["minimalist", "bold", "luxury", "playful"]:
+    variation = imagen.generate_images(
+        prompt=f"Product ad, {style} style, {brand_guidelines}",
+        number_of_images=1
+    )
+    variations.append(variation)
+```
+
+### 📚 Reference Documentation
+
+**Official Documentation:**
+- Vertex AI Multimodal: https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/overview
+- Gemini 2.5 Pro: https://cloud.google.com/vertex-ai/generative-ai/docs/models
+- Imagen 4: https://cloud.google.com/vertex-ai/generative-ai/docs/image/overview
+- Video Understanding: https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/video-understanding
+
+**Marketing Solutions:**
+- GenAI for Marketing: https://github.com/GoogleCloudPlatform/genai-for-marketing
+- ViGenAiR (video repurposing)
+- Adios (personalized ad images)
+
+**Pricing:**
+- Gemini 2.5 Pro: $3.50/1M input tokens, $10.50/1M output tokens
+- Imagen 4: $0.04/image
+- Video processing: Included in Gemini token pricing
+
+## When This Skill Activates
+
+This skill automatically activates when you mention:
+- Video processing, analysis, or understanding
+- Audio generation, music composition, or voiceovers
+- Image generation, ad creatives, or visual content
+- Marketing campaigns, content automation, or asset production
+- Gemini multimodal capabilities
+- Vertex AI media operations
+- Social media content, email marketing, or PMax campaigns
+
+## Integration with Other Tools
+
+**Google Cloud Services:**
+- Cloud Storage for media asset management
+- BigQuery for campaign analytics
+- Cloud Functions for automation triggers
+- Vertex AI Pipelines for content workflows
+
+**Third-Party Integrations:**
+- Social media APIs (LinkedIn, Twitter, Instagram)
+- Marketing automation platforms (HubSpot, Marketo)
+- CMS integrations (WordPress, Contentful)
+- DAM systems (Bynder, Cloudinary)
+
+## Success Metrics
+
+**Track These KPIs:**
+- Asset generation speed (baseline: 5 images/min)
+- Content approval rate (target: >80%)
+- Campaign personalization scale (target: 1000+ variants)
+- Cost per asset (target: <$0.10/image)
+- Time saved vs manual production (target: 90% reduction)
+
+---
+
+**This skill makes Jeremy a Vertex AI multimodal expert with instant access to video processing, audio generation, image creation, and marketing automation capabilities.**
--- a/skills/vertex-media-master/assets/README.md
+++ b/skills/vertex-media-master/assets/README.md
@@ -0,0 +1,26 @@
+# Skill Assets
+
+This directory contains static assets used by this skill.
+
+## Purpose
+
+Assets can include:
+- Configuration files (JSON, YAML)
+- Data files
+- Templates
+- Schemas
+- Test fixtures
+
+## Guidelines
+
+- Keep assets small and focused
+- Document asset purpose and format
+- Use standard file formats
+- Include schema validation where applicable
+
+## Common Asset Types
+
+- **config.json** - Configuration templates
+- **schema.json** - JSON schemas
+- **template.yaml** - YAML templates
+- **test-data.json** - Test fixtures
--- a/skills/vertex-media-master/references/README.md
+++ b/skills/vertex-media-master/references/README.md
@@ -0,0 +1,26 @@
+# Skill References
+
+This directory contains reference materials that enhance this skill's capabilities.
+
+## Purpose
+
+References can include:
+- Code examples
+- Style guides
+- Best practices documentation
+- Template files
+- Configuration examples
+
+## Guidelines
+
+- Keep references concise and actionable
+- Use markdown for documentation
+- Include clear examples
+- Link to external resources when appropriate
+
+## Types of References
+
+- **examples.md** - Usage examples
+- **style-guide.md** - Coding standards
+- **templates/** - Reusable templates
+- **patterns.md** - Design patterns
--- a/skills/vertex-media-master/scripts/README.md
+++ b/skills/vertex-media-master/scripts/README.md
@@ -0,0 +1,24 @@
+# Skill Scripts
+
+This directory contains optional helper scripts that support this skill's functionality.
+
+## Purpose
+
+Scripts here can be:
+- Referenced by the skill for automation
+- Used as examples for users
+- Executed during skill activation
+
+## Guidelines
+
+- All scripts should be well-documented
+- Include usage examples in comments
+- Make scripts executable (`chmod +x`)
+- Use `#!/bin/bash` or `#!/usr/bin/env python3` shebangs
+
+## Adding Scripts
+
+1. Create script file (e.g., `analyze.sh`, `process.py`)
+2. Add documentation header
+3. Make executable: `chmod +x script-name.sh`
+4. Test thoroughly before committing