AI Architect Newsletter

Purpose

Automated intelligence newsletter generation for AI Architects, combining latest AI news, FrankX research hub insights, and visual storytelling into production-ready HTML/text newsletters.

When to Use

• /ai-architect-newsletter

  - Generate weekly newsletter

• /ai-architect-newsletter daily

  - Quick daily brief

• /ai-architect-newsletter [topic]

  - Deep dive on specific topic

• /ai-architect-newsletter status

  - Review recent newsletters

Quick Start

Generate Weekly Newsletter

# Basic invocation (uses Opus 4.6 recommended)
/ai-architect-newsletter

# The system will:
# 1. Gather latest AI news (WebSearch)
# 2. Pull relevant research from /research hub
# 3. Generate visual newsletter (HTML + text)
# 4. Save to content/newsletters/YYYY-MM-DD-newsletter.html
# 5. Display summary with preview link

Newsletter Structure

1. Header Section

• Custom header image (AI Architect theme)
• Issue number, date, edition type
• Editor's note with Frank's perspective

2. Top Stories (5 key developments)

Each story includes:

• Headline with impact level
• "Why This Matters for AI Architects" analysis
• Visual asset (embedded or linked)
• Technical details or code examples
• Read more link to source

3. Deep Dive Section

One technical deep-dive featuring:

• Architecture diagrams
• Code examples
• Implementation patterns
• Production considerations

4. Research Spotlight

Highlight from FrankX research hub:

• Recent /research findings
• Cross-referenced with news
• Actionable insights
• Links to full research files

5. Tools Worth Watching

3-4 new tools with:

• Quick take (1-2 sentences)
• Use case for AI Architects
• GitHub/product link
• Integration considerations

6. Footer

• Newsletter archive link
• Social links (X, LinkedIn, GitHub)
• Unsubscribe/preferences
• Copyright notice

Intelligence Gathering

AI News Sources

See ai-news-sources.md for complete source list including:

Model Releases:

• OpenAI, Anthropic, Google DeepMind, Meta AI
• Cohere, Mistral, Stability AI, Hugging Face

Research:

• arXiv (cs.AI, cs.CL, cs.LG)
• Major lab blogs (OpenAI Research, Anthropic Research)
• Conference proceedings (NeurIPS, ICML, ACL)

Tools & Frameworks:

• GitHub trending (AI/ML repositories)
• LangChain, LlamaIndex, CrewAI releases
• MCP server ecosystem
• Agent frameworks (AutoGPT, BabyAGI updates)

Enterprise AI:

• Gartner AI reports
• Production AI case studies
• MLOps platform updates
• AI infrastructure news

Research Hub Integration

The skill automatically scans

/research/

directory for:

• Recent intelligence briefs (

  /research

  command output)
• Topic-specific deep dives
• Tool evaluations
• Pattern discoveries

Cross-references news with research to provide context and depth.

Visual Asset Strategy

AI Lab Logos:

• Store in

  assets/branding/labs/

• OpenAI, Anthropic, Google, Meta, etc.
• Used for story attribution

Generated Images:

• Nano Banana MCP for custom header images
• Architecture diagrams via Nano Banana
• AI-themed newsletter headers
• Concept illustrations for complex topics

Embedding vs Linking:

• Embed: Logo badges, small icons (<50KB)
• Link: Header images, diagrams, screenshots
• Optimization: All images compressed for email clients

Newsletter Modes

Weekly Newsletter (Default)

Trigger:

/ai-architect-newsletter

Output:

• 5-7 top stories
• 1 deep dive (1500-2000 words)
• Research spotlight
• Tools section
• ~15-20 min read time

Recommended Model: Opus 4.6 (deep synthesis)

Daily Brief

Trigger:

/ai-architect-newsletter daily

Output:

• Top 3 stories
• Quick takes (100-150 words each)
• Notable tools (1-2)
• ~5 min read time

Recommended Model: Sonnet 4.5 (faster)

Topic Deep Dive

Trigger:

/ai-architect-newsletter [topic]

Examples:

• /ai-architect-newsletter agentic systems

• /ai-architect-newsletter vision models

• /ai-architect-newsletter production RAG

Output:

• Focused on single topic
• Multi-source synthesis
• Code examples and patterns
• Architecture recommendations

Recommended Model: Opus 4.6 (depth)

Status Check

Trigger:

/ai-architect-newsletter status

Output:

• List recent newsletters
• Performance stats (if tracking enabled)
• Scheduled newsletters
• Archive links

Scripts Reference

gather-ai-news.ts

Purpose: Automated news gathering via WebSearch

Usage:

// Called automatically during newsletter generation
// Or manually: npx tsx scripts/gather-ai-news.ts --mode weekly

Output: JSON file with structured news data Location:

.claude/skills/ai-architect-newsletter/cache/news-YYYY-MM-DD.json

See script for configuration options.

process-research.ts

Purpose: Scan and index FrankX research hub

Usage:

// Scans /research/ directory for relevant content
// Indexes by topic, date, relevance

Output: Research index JSON Cross-reference: Matches news topics with research findings

generate-newsletter.ts

Purpose: Template engine for HTML/text output

Usage:

// Processes gathered data into newsletter format
// Applies templates, injects content, optimizes assets

Templates Used:

• assets/templates/newsletter.html

  - Main HTML template

• assets/templates/newsletter.txt

  - Plain text version

• assets/templates/social-preview.html

  - Social sharing card

optimize-images.ts

Purpose: Image processing and embedding

Usage:

// Compresses images for email clients
// Converts to base64 for embedding (small images)
// Uploads to CDN and generates links (large images)

Optimization:

• Resize for email clients (max 600px width)
• Compress JPEG (80% quality)
• PNG optimization via pngquant
• WebP fallbacks for modern clients

Voice & Style Guidelines

Frank's Newsletter Voice

See content-patterns.md for complete style guide.

Key Principles:

• Lead with impact: "Why this matters for AI Architects" first
• No hype: Precise technical language, factual analysis
• Actionable: Every story includes "What you can do" section
• Code over prose: Show implementations, not just concepts
• Humble expertise: Confident analysis without guru language

Tone:

• Studio energy at 2am (focused, building)
• Builder-to-builder communication
• Results over philosophy
• Specific numbers and examples

Avoid:

• Spiritual/guru language
• Grandiose claims ("revolutionary", "game-changing")
• Over-explaining fundamentals
• Marketing copy tone
• Emojis (unless data/code context)

Output Locations

FrankX Repository

/mnt/c/Users/Frank/FrankX/
├── content/newsletters/
│   ├── YYYY-MM-DD-weekly.html
│   ├── YYYY-MM-DD-weekly.txt
│   └── archive/
└── public/newsletters/
    └── images/
        └── YYYY-MM-DD/

AI Architect Academy Repository

/mnt/c/Users/Frank/ai-architect-academy/
├── newsletters/
│   ├── YYYY-MM-DD.html
│   └── archive/
└── assets/newsletter-images/

Multi-Repo Support

Repo Detection:

// Automatically detects current repo
const isAcademy = process.cwd().includes('ai-architect-academy');
const outputPath = isAcademy
  ? 'newsletters/'
  : 'content/newsletters/';

Cross-Repo Syncing:

• Newsletter generated in FrankX can be copied to Academy
• Research hub remains in FrankX only
• Academy uses external research links

Progressive Disclosure

This SKILL.md provides core workflow. For detailed information:

• ai-news-sources.md - Complete source list, search patterns, RSS feeds
• visual-standards.md - Design guidelines, brand colors, image specs
• content-patterns.md - Newsletter structures, voice guide, templates
• topic-clusters.md - AI Architect topic taxonomy, tagging system

Integration with Existing Commands

Synergy with /research:

# 1. Run research first
/research agentic systems

# 2. Generate newsletter with research integrated
/ai-architect-newsletter

# Research findings automatically included in newsletter

Synergy with /generate-social:

# After newsletter generation
/generate-social content/newsletters/YYYY-MM-DD-weekly.html

# Creates X thread + LinkedIn post from newsletter

Quality Checklist

Before sending newsletter:

• All sources verified and linked
• Code examples tested
• Images optimized (<500KB total)
• HTML renders in major email clients
• Plain text version readable
• Links functional
• Unsubscribe link present
• Mobile responsive
• No broken images
• Spell check passed

Model Recommendations

Opus 4.6 (Recommended for quality):

• Weekly newsletter generation
• Deep dive topics
• Multi-source synthesis
• Complex architecture diagrams

Sonnet 4.5 (Faster):

• Daily briefs
• Status checks
• Quick updates
• Research indexing

When to use each:

• First newsletter run: Opus 4.6 (establish quality baseline)
• Daily operations: Sonnet 4.5 (speed)
• Important editions: Opus 4.6 (quality)
• Routine updates: Sonnet 4.5 (efficiency)

Troubleshooting

Newsletter not generating:

• Check WebSearch tool availability
• Verify /research directory accessible
• Ensure write permissions to output directory

Missing images:

• Verify Nano Banana MCP connection
• Check image optimization script
• Ensure sufficient disk space

Research not integrated:

• Run

  /research

  command first to generate content
• Verify research files in /research/ directory
• Check file permissions

Multi-repo issues:

• Verify current working directory
• Check path configurations in scripts
• Ensure both repos accessible

Future Enhancements

• Automated scheduling (weekly cron)
• Email distribution integration (Resend MCP)
• Analytics tracking (open rates, clicks)
• A/B testing for subject lines
• Subscriber segmentation
• RSS feed generation
• Archive website

---

Created: 2026-02-16 Status: Production Ready Model: Works with Sonnet 4.5, Opus 4.6 recommended for quality Line Count: 458 (under 500-line rule ✅)