Bloomberg Daily Briefing

Automated daily news delivery from Bloomberg, optimized for e-ink readers.

OPDS Feed

https://mylesmcook.github.io/bloomberg-daily/opds.xml

Add this URL to your e-reader's OPDS browser to automatically sync daily briefings.

Features

• Automated daily fetch - Runs weekdays at 6:00 AM CST
• E-ink optimized - Custom CSS and fonts (Newsreader) for readable typography
• Curated sections - AI, Technology, Industries, Latest
• Rolling archive - Last 7 issues always available
• Zero maintenance - Fully automated via GitHub Actions + Pages
• Dark mode support - Automatic theme switching via CSS media queries
• Health monitoring - JSON endpoint for system status
• Debug mode - Verbose logging on demand
• Build diagnostics - Metadata embedded in each EPUB

Architecture

┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  GitHub Actions │────▶│  GitHub Pages    │────▶│  E-ink Reader   │
│  (6am weekdays) │     │  (Static OPDS)   │     │  (CrossPoint)   │
└─────────────────┘     └──────────────────┘     └─────────────────┘
        │                        │
        │                        │
        ▼                        ▼
   Fetch & Process          Serve EPUBs
   Bloomberg News           via OPDS 1.2

Files

File	Purpose
bloomberg_filtered.recipe	Calibre recipe for fetching Bloomberg
process_epub.py	Post-processor for CSS/fonts/cleanup
generate_opds.py	OPDS catalog generator
cleanup_old_books.py	Maintains 7-day rolling archive
stylesheet.css	E-ink optimized styles with dark mode
fonts/	Newsreader font family (Google Fonts)
books/	EPUB archive (auto-managed)
opds.xml	Generated OPDS catalog
health.json	System health status endpoint

Manual Trigger

To manually fetch today's news:

1. Go to Actions tab
2. Select Fetch Bloomberg Daily
3. Click Run workflow
4. Optionally enable Debug mode for verbose logging
5. Wait ~2 minutes for fetch + deploy

Monitoring & Health

Health Check Endpoint

https://mylesmcook.github.io/bloomberg-daily/health.json

Returns system status including book count, sizes, and dates:

{
  "status": "ok",
  "book_count": 5,
  "newest_book": "2026-01-31",
  "total_size_mb": 18.5
}

Diagnostic Manifest

Each EPUB contains _diagnostics.json with build metadata:

• workflow_run_id - GitHub Actions run ID
• git_sha - Commit that built this EPUB
• build_time - When the EPUB was created
• article_count - Number of articles processed

Extract with: unzip -p Bloomberg_*.epub _diagnostics.json

Troubleshooting

Debug Mode

Enable verbose logging for any workflow run:

1. Go to Actions → Run workflow
2. Check Enable debug mode
3. View detailed logs in the run output

Common Issues

Symptom	Check	Solution
No new EPUB	health.json last_update	Check workflow run logs
Empty EPUB	Workflow logs for "VALIDATION FAILED"	Bloomberg may be down
Workflow timeout	Calibre fetch step	Bloomberg rate limiting
Missing articles	Recipe sections list	Update ALLOWED_SECTIONS

Debug Artifacts

On workflow failure, debug artifacts are automatically uploaded:

• temp_output/calibre.log - Calibre fetch output
• temp_output/process.log - EPUB processing output
• temp_output/opds.log - Catalog generation output

Find them in the workflow run under Artifacts.

Sections Included

• AI - Artificial intelligence and machine learning news
• Technology - Tech industry coverage
• Industries - Business and sector news
• Latest - Breaking news and updates

Device Setup (CrossPoint OS)

1. Open OPDS Browser on your device
2. Add new catalog with URL: https://mylesmcook.github.io/bloomberg-daily/opds.xml
3. No authentication required
4. Set auto-sync schedule as desired

Configuration

Change sections

Edit bloomberg_filtered.recipe:

ALLOWED_SECTIONS = ['ai', 'technology', 'industries', 'latest']

Change schedule

Edit .github/workflows/fetch-bloomberg.yml:

schedule:
  - cron: '0 12 * * 1-5'  # 12:00 UTC = 6:00 AM CST, Mon-Fri

Change archive size

Edit the workflow or cleanup_old_books.py:

python cleanup_old_books.py --keep 7  # Keep last 7 issues

Local Development

# Install Calibre first (https://calibre-ebook.com)

# Fetch and process
ebook-convert bloomberg_filtered.recipe output/Bloomberg_Raw.epub --output-profile=generic_eink_hd
python process_epub.py output/Bloomberg_Raw.epub books/Bloomberg_$(date +%Y-%m-%d).epub

# Generate catalog
python generate_opds.py

---

Powered by Calibre, GitHub Actions, and GitHub Pages

For personal use only. Bloomberg content © Bloomberg L.P.