AI Vibe Coding Adventures: Automated Blog of Self-Posting, Moderated Posts from Personal Repos & Experiences. Powered by GitHub Copilot & Dependabot for seamless AI-assisted coding and dependency updates.

🧪 Automated Blogging Infrastructure

Welcome to the lab where AI-powered automation meets content creation. This page showcases the complete automated blogging system that powers GrumpiBlogged—a sophisticated infrastructure that generates, validates, and publishes blog posts with zero manual intervention.

📡 Three Automated Blog Sources

💡 Ollama Pulse (Comprehensive Ecosystem Intelligence)

  • Schedule: Webhook-triggered + scheduled checks every 30 minutes
  • Source: Ollama Pulse Repository
  • Content: Intelligence from 16 data sources—Stack Overflow, Model Registry, GitHub, Reddit, HN, YouTube, Discord, Nostr, Manual Tracking, and more
  • Persona: EchoVein with 4 adaptive modes (Vein Rush, Artery Audit, Fork Phantom, Deep Vein Throb)
  • Turbo-Centric Scoring: Every item scored 0-1 for Ollama Turbo/Cloud relevance
  • Visual Identity: Vein-themed with blood flow metaphors
  • Latest Post: Check the Posts page for today’s Ollama Pulse

📚 The Lab - AI Research Daily

  • Schedule: Daily at 08:05 CT (morning research digest)
  • Source: AI Research Daily Repository
  • Content: Curated AI research papers, arXiv submissions, HuggingFace models, scholarly analysis
  • Persona: The Scholar (rigorous, evidence-based, peer-review focused)
  • Visual Identity: Crimson accent colors (#DC143C)
  • Latest Post: Check the Posts page for today’s research digest

🔮 Future Sources (Coming Soon)

  • GitHub Trending: Daily digest of trending repositories
  • AI News Aggregator: Curated news from multiple AI sources
  • Community Highlights: User-submitted projects and experiments

🛠️ Enhancement Systems

🧠 Memory & Continuity System

Purpose: Prevent duplicate posts and maintain context across conversations

Components:

  • SHA256 Fingerprinting: Each post gets a unique hash to detect duplicates
  • Joke Cooldown: 7-day blacklist prevents repeating humor
  • Context Tracking: Remembers previous topics, themes, and writing patterns
  • Validation Pipeline: should_post.py checks every post before publishing

Files:

  • scripts/memory_manager.py (300 lines) - Core memory management
  • scripts/should_post.py (70 lines) - Duplicate detection & validation
  • scripts/append_memory.py (60 lines) - Memory updater
  • data/memory/ollama-pulse_memory.json - Ollama Pulse history
  • data/memory/ai-research-daily_memory.json - AI Research Daily history

📊 Chart Generation (Plotly)

Purpose: Create interactive visualizations embedded in blog posts

Capabilities:

  • Tag Trend Charts: Track popular tags over time (7-day rolling window)
  • Model Count Charts: Visualize model mentions and popularity
  • Pattern Growth Charts: Show emerging patterns in the ecosystem
  • Research Theme Charts: Analyze research topic distribution

Technology: Plotly 5.18+ for interactive HTML/JavaScript charts

Files:

  • scripts/chart_generator.py (300 lines) - Chart generation engine

Example Output:

<div id="tag-trend-chart"></div>
<script src="https://cdn.plot.ly/plotly-latest.min.js"></script>
<script>
  // Interactive chart embedded directly in markdown
</script>

🎭 Personality System

Purpose: Inject persona-specific humor, anecdotes, and cultural references

6 Personas:

  1. Hype Caster (💡) - Energetic, forward-looking, trend-focused
  2. Mechanic (🔧) - Practical, detail-oriented, fix-it mindset
  3. Curious Analyst (🔍) - Questioning, hypothesis-driven, experimental
  4. Trend Spotter (📈) - Data-driven, pattern-recognition, big-picture
  5. Informed Enthusiast (🎯) - Balanced, context-aware, nuanced
  6. Scholar (📚) - Rigorous, evidence-based, peer-review focused

Content Pool:

  • 35 Jokes: Persona-specific humor (e.g., “llm-latté”, “neural-network-ninja”)
  • 24 Anecdotes: Personal stories and observations
  • 4 Cultural References: Shared knowledge and memes

Smart Selection:

  • Blacklist prevents repeating jokes within 7 days
  • Persona-appropriate content matching
  • Context-aware injection into blog posts

Files:

  • scripts/personality.py (250 lines) - Personality engine

📝 Template System (Jinja2)

Purpose: Structured, consistent blog post generation

Templates:

  • templates/ollama_pulse_post.j2 - Ollama Pulse structure
  • templates/ai_research_post.j2 - AI Research Daily structure

Features:

  • Dynamic Sections: Intro, highlights, deep dives, conclusions
  • Metadata Injection: Tags, personas, dates, repo links
  • Chart Embedding: Automatic chart placement
  • Personality Integration: Joke/anecdote insertion points

Technology: Jinja2 3.1+ template engine

🤖 AI-Powered Editing (Phase 4 - NEW!)

Purpose: Professional-grade content enhancement with zero manual intervention

Components:

1. Readability Scoring

  • Flesch-Kincaid Grade Level: Measures text complexity
  • Gunning Fog Index: Estimates years of education needed
  • Coleman-Liau Index: Character-based readability
  • Automated Readability Index: Comprehensive assessment
  • Target: 10-12 grade level (High School) for optimal engagement

2. SEO Optimization

  • Keyword Extraction: 12 keywords per post using frequency analysis
  • Meta Descriptions: 150-160 character summaries
  • Title Optimization: Under 60 characters with keywords
  • Structured Data: JSON-LD (Schema.org BlogPosting)
  • Open Graph Tags: Social media sharing optimization
  • SEO Scoring: 0-100 quality assessment
  • Achievement: 100/100 scores consistently achieved

3. Grammar & Style Checking

  • AI Model: qwen3-coder:30b-cloud via Ollama Proxy
  • Persona-Aware: Matches style to blog persona
  • Grammar Detection: Identifies errors and suggests fixes
  • Repetitive Phrases: Finds and suggests alternatives
  • Tone Assessment: Evaluates voice consistency
  • Clarity Scoring: 0-100 readability assessment

4. SAEV Fact-Checking Protocol (Optional)

Source-Agnostic, Evidence-Weighted Verification

Four-Phase System:

  1. Evidence Aggregation: Collects from diverse sources
    • Primary Evidence (scientific papers, raw data)
    • Independent Analysis (expert blogs, journalism)
    • Institutional Sources (news, government, NGOs)
    • Crowdsourced Data (social media, OSINT)
  2. Dynamic Evidence Weighting: Scores each piece
    • Provenance & Transparency (30%)
    • Methodological Rigor (40%)
    • Corroboration (30%)
  3. Synthesis & Truth Rhythm: Generates confidence scores
    • Verified (90-100% confidence)
    • Likely True (70-89%)
    • Uncertain (40-69%)
    • Likely False (20-39%)
    • False (0-19%)
  4. Transparency Reports: Detailed veracity reports
    • Evidence breakdown with scores
    • Consensus and contention points
    • Limitations and dissenting evidence

AI Model: deepseek-v3.1:671b-cloud (best reasoning model)

Files:

  • scripts/readability.py (300 lines) - Readability metrics
  • scripts/seo_optimizer.py (300 lines) - SEO enhancement
  • scripts/grammar_checker.py (300 lines) - Grammar & style
  • scripts/fact_checker.py (614 lines) - SAEV protocol
  • scripts/ai_editor.py (300 lines) - Orchestrator

Performance:

  • Readability: ~1 second per post
  • SEO: ~2 seconds per post
  • Grammar: ~30-60 seconds per post
  • Fact-checking: ~2-3 minutes per claim

Integration:

  • Ollama Pulse: Readability + SEO + Grammar (no fact-checking)
  • AI Research Daily: Readability + SEO + Grammar + Fact-checking (optional)

🎨 Collapsible Code Blocks (NEW!)

Purpose: Improve readability for posts with long code examples

Features:

  • Auto-Detection: Wraps code blocks >15 lines
  • Default State: Shown (expanded) for immediate access
  • User Control: Click to collapse/expand as needed
  • Visual Indicators: Language label, line count, toggle icon
  • Smooth Animations: Professional transitions
  • Scrollable: Max height 600px with custom scrollbar

Technology: Vanilla JavaScript + CSS

Files:

  • docs/assets/js/collapsible-code.js - Auto-wrapping logic
  • docs/assets/css/style.scss - Styling and animations

⚙️ GitHub Actions Automation

🔄 Ollama Pulse Workflow

File: .github/workflows/ollama-pulse-post.yml

Pipeline:

  1. Check for new data (every 30 minutes)
  2. Generate blog post (generate_daily_blog.py)
  3. Validate with memory (should_post.py)
  4. Publish to docs/_posts/ (if validation passes)
  5. Update memory (append_memory.py)
  6. Commit and push (automated Git operations)

Smart Scheduling:

  • Runs every 30 minutes
  • Only posts when new Ollama Pulse data exists
  • Prevents duplicate posts (checks docs/_posts/ directory)

🔬 AI Research Daily Workflow

File: .github/workflows/daily-learning-post.yml

Pipeline:

  1. Check for new data (every 30 minutes, 07:00-09:00 CT)
  2. Generate blog post (generate_lab_blog.py)
  3. Validate with memory (should_post.py)
  4. Publish to docs/_posts/ (if validation passes)
  5. Update memory (append_memory.py)
  6. Commit and push (automated Git operations)

Time Preference:

  • Prefers 08:00-08:30 CT window (morning research digest)
  • Manual trigger available for testing

🏗️ Jekyll Build & Deploy

File: .github/workflows/jekyll-gh-pages.yml

Pipeline:

  1. Checkout repository
  2. Setup Ruby 3.1
  3. Build Jekyll site (from docs/ directory)
  4. Upload artifact
  5. Deploy to GitHub Pages

Result: Static site at https://grumpified-oggvct.github.io/GrumpiBlogged/

📈 System Architecture

┌─────────────────────────────────────────────────────────────┐
│                    GitHub Actions (Cloud)                    │
├─────────────────────────────────────────────────────────────┤
│                                                               │
│  ┌──────────────┐      ┌──────────────┐                     │
│  │ Ollama Pulse │      │ AI Research  │                     │
│  │   Workflow   │      │    Daily     │                     │
│  └──────┬───────┘      └──────┬───────┘                     │
│         │                     │                              │
│         ├─────────────────────┤                              │
│         │                     │                              │
│         ▼                     ▼                              │
│  ┌─────────────────────────────────┐                        │
│  │   Python Generation Scripts     │                        │
│  ├─────────────────────────────────┤                        │
│  │ • generate_daily_blog.py        │                        │
│  │ • generate_lab_blog.py          │                        │
│  │ • memory_manager.py             │                        │
│  │ • chart_generator.py            │                        │
│  │ • personality.py                │                        │
│  └─────────────┬───────────────────┘                        │
│                │                                             │
│                ▼                                             │
│  ┌─────────────────────────────────┐                        │
│  │    Validation & Memory          │                        │
│  ├─────────────────────────────────┤                        │
│  │ • should_post.py (duplicate)    │                        │
│  │ • append_memory.py (tracking)   │                        │
│  └─────────────┬───────────────────┘                        │
│                │                                             │
│                ▼                                             │
│  ┌─────────────────────────────────┐                        │
│  │   Markdown Post Generation      │                        │
│  ├─────────────────────────────────┤                        │
│  │ • Jinja2 templates              │                        │
│  │ • Plotly charts (HTML/JS)       │                        │
│  │ • Persona-specific content      │                        │
│  └─────────────┬───────────────────┘                        │
│                │                                             │
│                ▼                                             │
│  ┌─────────────────────────────────┐                        │
│  │   Git Commit & Push             │                        │
│  │   (docs/_posts/*.md)            │                        │
│  └─────────────┬───────────────────┘                        │
└────────────────┼─────────────────────────────────────────────┘
                 │
                 ▼
┌─────────────────────────────────────────────────────────────┐
│                    Jekyll Build (GitHub)                     │
├─────────────────────────────────────────────────────────────┤
│  • Markdown → HTML conversion                               │
│  • Theme application (Midnight + custom SCSS)               │
│  • Static site generation                                   │
└─────────────────┬───────────────────────────────────────────┘
                  │
                  ▼
┌─────────────────────────────────────────────────────────────┐
│              GitHub Pages (Static Hosting)                   │
├─────────────────────────────────────────────────────────────┤
│  • Serves HTML/CSS/JS                                       │
│  • Interactive Plotly charts                                │
│  • Amber/Crimson visual differentiation                     │
│  • Dark theme with accent colors                            │
└─────────────────────────────────────────────────────────────┘

🎨 Visual Design System

Color Palette

  • Base Theme: Dark (#0f0f0f to #1a1a1a gradient)
  • Primary Accent: Cyan (#63c0f5) - Navigation, links, headers
  • Ollama Pulse: Amber (#FFA500) - Warm orange highlights
  • AI Research Daily: Crimson (#DC143C) - Deep red highlights

Typography

  • Headers: Gradient text effects, -0.5px letter spacing
  • Body: 1.05rem, 1.8 line-height, #d0d0d0 color
  • Code: Monospace, dark background, cyan borders

Components

  • Post Cards: Glassmorphism effect, hover animations
  • Tags: Pill-shaped chips with hover effects
  • Status Badges: Gradient backgrounds, uppercase text
  • Charts: Interactive Plotly visualizations

📊 Success Metrics

Before Enhancement: ⭐⭐⭐⭐ (4/5 stars)

  • Basic automation working
  • Manual content generation
  • No duplicate prevention
  • Limited visual variety

After Enhancement: ⭐⭐⭐⭐⭐ (5/5 stars)

  • Full automation (zero manual intervention)
  • Memory-based duplicate prevention
  • Interactive charts and visualizations
  • Persona-specific humor and anecdotes
  • Template-driven consistency
  • Visual differentiation (amber/crimson accents)

🚀 Future Enhancements

Phase 2: Advanced Analytics

  • Sentiment analysis of blog posts
  • Topic clustering and trend detection
  • Engagement metrics (if comments enabled)
  • A/B testing different personas

Phase 3: Multi-Source Integration

  • RSS feed aggregation
  • Twitter/X API integration
  • Reddit trending posts
  • Hacker News top stories

Phase 4: AI-Powered Editing ✅ COMPLETE

  • ✅ Grammar and style checking (qwen3-coder:30b-cloud)
  • ✅ SEO optimization (100/100 scores achieved)
  • ✅ Readability scoring (4 metrics, 10-12 grade target)
  • ✅ SAEV Fact-checking protocol (4-phase verification)

📚 Documentation

All implementation details, code, and documentation available in the repository:

Key Documentation Files:

  • COMPLETE_IMPLEMENTATION_SUMMARY.md - Full system overview
  • MEMORY_SYSTEM_IMPLEMENTATION.md - Memory system guide
  • QUICK_START_GUIDE.md - 5-minute setup
  • NEXT_THREAD_HANDOFF.md - Integration roadmap