commit 0b3ddcd76a237d3a31dbff16aa29e1d22d288ccc Author: Zhongwei Li Date: Sun Nov 30 08:21:41 2025 +0800 Initial commit diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json new file mode 100644 index 0000000..9b4e716 --- /dev/null +++ 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" + ] +} \ No newline at end of file diff --git a/README.md b/README.md new file mode 100644 index 0000000..590753c --- /dev/null +++ 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. diff --git a/commands/vertex-campaign.md b/commands/vertex-campaign.md new file mode 100644 index 0000000..446e7c1 --- /dev/null +++ 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: ~$5-7 per complete campaign** + +## Best Practices + +1. **Brand Consistency**: Provide brand guidelines in prompt +2. **Batch Generation**: Generate multiple variations simultaneously +3. **Quality Control**: Review and iterate on generated assets +4. **Version Control**: Save prompts and outputs for reproducibility +5. **A/B Testing**: Generate 3-5 variations of each asset + +## Integration with Marketing Stack + +**Export to:** +- Google Ads (PMax campaigns) +- Meta Business Suite (Facebook/Instagram) +- LinkedIn Campaign Manager +- Email marketing platforms (HubSpot, Mailchimp) +- CMS platforms (WordPress, Contentful) + +## Performance Tracking + +**Monitor:** +- Asset generation time +- Cost per asset +- Approval rates +- Campaign performance metrics +- ROI vs traditional production + +--- + +**This command turns Jeremy into a one-person marketing agency powered by Vertex AI multimodal capabilities.** diff --git a/plugin.lock.json b/plugin.lock.json new file mode 100644 index 0000000..31859a6 --- /dev/null +++ b/plugin.lock.json @@ -0,0 +1,61 @@ +{ + "$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": [] + } +} \ No newline at end of file diff --git a/skills/vertex-media-master/SKILL.md b/skills/vertex-media-master/SKILL.md new file mode 100644 index 0000000..27ecf33 --- /dev/null +++ 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.** diff --git a/skills/vertex-media-master/assets/README.md b/skills/vertex-media-master/assets/README.md new file mode 100644 index 0000000..148d1c3 --- /dev/null +++ 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 diff --git a/skills/vertex-media-master/references/README.md b/skills/vertex-media-master/references/README.md new file mode 100644 index 0000000..bf97184 --- /dev/null +++ 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 diff --git a/skills/vertex-media-master/scripts/README.md b/skills/vertex-media-master/scripts/README.md new file mode 100644 index 0000000..9f9c7ad --- /dev/null +++ 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