Automatic Publication to Bluesky

This project uses a CI workflow to automatically publish bioinformatics blog posts from selected RSS feeds to a Bluesky account every day at 09:00 UTC. Each run scans for posts published within the previous 24 hours and shares them on Bluesky with rich link previews, ensuring no duplicates are posted and avoiding the need to track previously published entries.

The use case here is bioinfoblogs.

Setup

1. Install dependencies

pip install atproto pyyaml

Or add to your pyproject.toml:

dependencies = [
    "feedparser>=6.0.12",
    "jinja2>=3.1.6",
    "atproto>=0.0.55",
    "pyyaml>=6.0",
]

2. Create a Bluesky app password

1. Log in to Bluesky
2. Go to Settings → App Passwords
3. Create a new app password
4. Save it securely

3. Configure credentials in GitHub

1. Go to your repository on GitHub
2. Navigate to Settings → Secrets and variables → Actions
3. Click New repository secret
4. Add two secrets:
   • Name: BLUESKY_USERNAME
     Value: your-handle.bsky.social (without the @)
   • Name: BLUESKY_PASSWORD
     Value: Your Bluesky app password created in step 2

These secrets will be securely used by the GitHub Actions workflow.

Usage

Test mode (dry-run)

To see what would be published without actually posting:

python bluesky_publisher.py \
  --username your-handle.bsky.social \
  --password your-app-password \
  --dry-run

Real publication

python bluesky_publisher.py \
  --username your-handle.bsky.social \
  --password your-app-password

Options

• --hours N: Fetch posts from the last N hours (default: 24)
• --dry-run: Test mode without actual publication

Automation with GitHub Actions

Create .github/workflows/bluesky.yml:

name: Publish to Bluesky

on:
  schedule:
    # Run every day at 9 a.m. UTC
    - cron: '0 9 * * *'
  workflow_dispatch: # Allows manual execution
    inputs:
      dry_run:
        description: 'Dry-run - does not actually publish'
        required: false
        type: boolean
        default: false

jobs:
  publish:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      
      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.13'
      
      - name: Install dependencies
        run: |
          pip install feedparser atproto pyyaml
      
      - name: Publish to Bluesky
        env:
          BLUESKY_USERNAME: ${{ secrets.BLUESKY_USERNAME }}
          BLUESKY_PASSWORD: ${{ secrets.BLUESKY_PASSWORD }}
        run: |
          python bluesky_publisher.py \
            --username "$BLUESKY_USERNAME" \
            --password "$BLUESKY_PASSWORD" \
            --hours 24 \
            ${{ github.event.inputs.dry_run == 'true' && '--dry-run' || '' }}

How it works

1. The script reads RSS feeds from feeds.yaml
2. It fetches only posts published in the last 24 hours (configurable with --hours)
3. It publishes each post on Bluesky with:
   • The article title
   • The source (blog name)
   • A rich link to the article
4. It saves published URLs to avoid duplicates

Post format

Example of a generated post:

📝 A new method for genome assembly
✍️ Dave Tang's blog
🔗 https://davetang.org/muse/...

RSS Feed Management

RSS feeds are defined in feeds.yaml:

feeds:
  - name: "Blog name"
    url: "https://example.com/feed.xml"
  
  - name: "Another blog"
    url: "https://example2.com/rss"

To add a new feed, simply edit this file.

Tips

• Always use --dry-run first to check what will be published
• Adjust --hours according to your CI execution frequency
• If CI runs every 24h, use --hours 24 to avoid duplicates