🧹 blueskydelete

🔄 Recreating some of the functionality of https://tweetdelete.net/, but for Bluesky!

✨ What it does: Automatically cleans up your Bluesky posts, reposts, and likes older than a configurable number of days (default: 90 days).

🎯 Goal: Keep your feed fresh by only maintaining recent content while cleaning up the old stuff.

✨ Features

• 🎨 Colored logging - Color-coded log levels for better output readability
• 🔧 Configurable log levels - Set LOG_LEVEL to control verbosity (DEBUG, INFO, WARNING, ERROR)
• 🛡️ Dry run mode - Safe testing without actually deleting content
• 📅 Flexible date ranges - Configure how far back to delete with DAYS_AGO
• 📝 Multiple content types - Handles posts, reposts, and likes

🚀 Quick Start (macOS)

1. 📥 Clone this repo
2. 🐍 Check Python version - See .python-version file (other versions might work)
3. 🏠 Create virtual environment: python3 -m venv .venv
4. ⚡ Activate environment: source .venv/bin/activate
5. 📦 Install dependencies:
   • Development: pip install -e .'[dev]'
   • Production: pip install -e .
6. ⚙️ Setup environment: touch .env & cp .env.sample .env then edit .env with your values
7. 🎬 Run script: python main.py

⚙️ Environment Variables

Variable	Description	Default
USERNAME	👤 Your Bluesky username/handle	-
PASSWORD	🔐 Your Bluesky app password (not main password!)	-
DRY_RUN	🛡️ Safe mode: true for testing, false for real deletion	true
DAYS_AGO	📅 How many days back to delete content	90
LOG_LEVEL	📊 Logging detail: DEBUG, INFO, WARNING, ERROR	INFO

💡 Usage Examples

# Normal run (dry run mode, INFO logging)
python main.py

# Actually delete content with debug logging
LOG_LEVEL=DEBUG DRY_RUN=false python main.py

# Delete content older than 30 days
DAYS_AGO=30 DRY_RUN=false python main.py

# Quiet mode - only show warnings and errors
LOG_LEVEL=WARNING python main.py

🛡️ Safety Features

• 🔒 Dry run by default - Script runs in safe mode unless explicitly disabled
• ⚠️ Colored warnings - Dry run operations show in yellow to make it clear no actual deletion is happening
• 🔍 Detailed logging - Use LOG_LEVEL=DEBUG to see exactly what would be deleted before running for real

🤖 Automated Scheduling

The repository includes a GitHub Actions workflow (.github/workflows/cleanup-feed.yml) that automatically runs the cleanup script on a schedule.

⏰ Current schedule: Every Friday at 5:00 AM UTC (0 5 * * 5)

🔧 Setting up automated runs:

1. 📊 Repository Variables (Settings → Secrets and variables → Actions → Variables):

   • USERNAME - 👤 Your Bluesky username/handle
   • DRY_RUN - 🛡️ Set to false to actually delete, true for dry run
   • DAYS_AGO - 📅 Number of days back to delete (optional, defaults to 90)

2. 🔐 Repository Secrets (Settings → Secrets and variables → Actions → Secrets):

   • PASSWORD - 🔑 Your Bluesky app password

3. ▶️ Manual runs: You can also trigger the workflow manually from the Actions tab

⏰ Customizing the schedule:

Edit the cron expression in .github/workflows/cleanup-feed.yml:

schedule:
  - cron: "0 5 * * 5" # Every Friday at 5 AM UTC

📅 Common cron patterns:

• 0 0 * * 0 - 🗓️ Weekly on Sunday at midnight
• 0 12 1 * * - 📆 Monthly on the 1st at noon
• 0 6 * * 1,3,5 - 📋 Monday, Wednesday, Friday at 6 AM

💡 Note: The workflow uses colored logging output which displays nicely in GitHub Actions logs, making it seamless to monitor the cleanup process.