neo-notebook

Setup Guide

Complete installation and configuration guide for the AI Security Intelligence Engine.

Prerequisites

• Python 3.11+
• Git
• (Optional) Ollama for LLM summarization

Installation

1. Clone Repository

git clone <repository-url>
cd neo-notebook.github.io

2. Create Virtual Environment

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

pip install --upgrade pip
pip install -r requirements.txt

4. Configure Sources

Edit config/sources.yaml to add your RSS feeds:

sources:
  - name: "Your Source Name"
    type: "rss"
    url: "https://example.com/feed.xml"
    category: "security_research"
    credibility_tier: "medium"
    enabled: true

Running the Pipeline

Daily Pipeline (Local)

python src/main.py --mode daily

This will:

1. Fetch content from configured sources
2. Process, score, and rank items
3. Summarize top items
4. Generate public brief and private archive

Weekly Pipeline (Local)

python -c "
from src.outputs.json_generator import JSONGenerator
from src.outputs.pdf_generator import PDFGenerator
import json, glob

# Load last 30 days
items = []
for f in glob.glob('private/feeds/*.json'):
    with open(f) as file:
        items.extend(json.load(file)['items'])

# Generate weekly outputs
JSONGenerator().generate_trends(items, 'docs/public/data/trends_30d.json')
PDFGenerator().generate_weekly_brief(items[:50], 'private/artifacts/weekly_brief.pdf')
"

GitHub Actions Setup

1. Enable GitHub Pages

• Go to repository Settings → Pages
• Source: Deploy from branch main
• Folder: /docs/public

2. Set Secrets (Optional)

If using cloud LLM instead of Ollama:

• Go to Settings → Secrets → Actions
• Add secret: OLLAMA_URL or LLM_API_KEY

3. Enable Workflows

GitHub Actions will automatically run:

• Daily: 7:00 AM America/Edmonton (13:00 UTC)
• Weekly: Sunday 9:00 AM America/Edmonton (15:00 UTC)

Optional: Ollama Setup

For local LLM summarization:

# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Pull model
ollama pull llama3.2

# Start Ollama (runs on localhost:11434)
ollama serve

The pipeline will automatically use Ollama if available, otherwise falls back to metadata-based summaries.

Customization

Adjust Scoring Weights

Edit config/scoring_weights.yaml:

dimensions:
  relevance:
    weight: 0.35  # Adjust to prioritize different dimensions

Add/Remove Sources

Edit config/sources.yaml and set enabled: true/false

Modify Tags

Edit config/tags_taxonomy.yaml to add custom tags

Troubleshooting

No items fetched

• Check RSS feed URLs are valid
• Verify enabled: true in sources.yaml
• Test feeds manually: curl <feed-url>

Tests failing

PYTHONPATH=. pytest tests/ -v

GitHub Actions not running

• Check workflow files in .github/workflows/
• Verify Actions are enabled in repository settings

Directory Structure

├── .github/workflows/      # GitHub Actions
├── config/                 # Configuration files
├── docs/                   # Documentation
│   └── public/            # GitHub Pages site
├── private/               # Private outputs (gitignored)
│   ├── feeds/            # Daily archives
│   └── artifacts/        # Weekly PDFs, drafts
├── src/                   # Python source code
└── tests/                 # Test suite

Next Steps

1. Add real RSS feeds to config/sources.yaml
2. Run daily pipeline locally to verify
3. Push to GitHub to enable automation
4. Monitor GitHub Pages site at https://<username>.github.io