Skip to main content
Your repr stories are valuable. They represent months (or years) of work you’ve shipped. You don’t want to lose them. Good news: repr stores everything as plain JSON files in ~/.repr/. But having an explicit backup workflow is still smart. Here’s how.

​
Quick Backup

Want a backup right now? One command: 
repr data backup --output ~/backups/repr-backup-$(date +%Y-%m-%d).json
Output: 
Creating backup...

✓ Stories: 45 files (2.1 MB)
✓ Profiles: 3 files (340 KB)
✓ Config: 1 file (8 KB)

Backup saved to:
  ~/backups/repr-backup-2026-01-05.json

Total size: 2.5 MB
Done. You now have a complete backup of all your repr data in a single JSON file.

​
What Gets Backed Up?

The backup includes:
  • ✅ All stories (titles, narratives, metadata, featured/hidden status)
  • ✅ Profile settings (bio, location, availability)
  • ✅ Configuration (LLM settings, tracked repos, hooks)
  • ✅ Queue (pending commits waiting for generation)
What doesn’t get backed up:
  • ❌ Your git repositories (use git for that)
  • ❌ Your API keys (stored in OS keychain, not in repr)
  • ❌ Auth tokens (you’ll re-login after restore)

​
Where to Store Backups

Some good options: 
# Backup to Dropbox
repr data backup --output ~/Dropbox/repr-backup.json

# Backup to Google Drive
repr data backup --output ~/Google\ Drive/repr-backup.json

# Backup to iCloud
repr data backup --output ~/Library/Mobile\ Documents/com~apple~CloudDocs/repr-backup.json
Now your backup syncs automatically across devices and stays safe in the cloud.

​
2. External Drive

repr data backup --output /Volumes/Backup/repr-backup-$(date +%Y-%m-%d).json
Good for air-gapped systems or if you don’t trust cloud storage.

​
3. Version Control

# Backup to a git repo
repr data backup --output ~/backup-repo/repr-$(date +%Y-%m-%d).json
cd ~/backup-repo
git add .
git commit -m "Backup $(date +%Y-%m-%d)"
git push
Gives you version history of your backups. Useful if you want to roll back to an earlier state.

​
Automated Backups

Don’t rely on remembering to run backups. Automate it.

​
Weekly Backup (macOS)

Create a launchd service: 
# Create the backup script
cat > ~/scripts/repr-backup.sh << 'EOF'
#!/bin/bash
BACKUP_DIR=~/Dropbox/repr-backups
mkdir -p "$BACKUP_DIR"
repr data backup --output "$BACKUP_DIR/repr-$(date +%Y-%m-%d).json"
# Keep only last 10 backups
ls -t "$BACKUP_DIR"/repr-*.json | tail -n +11 | xargs rm -f
EOF

chmod +x ~/scripts/repr-backup.sh

# Test it
~/scripts/repr-backup.sh
Then set up a weekly cron job: 
crontab -e

# Add this line (runs every Sunday at 8pm)
0 20 * * 0 ~/scripts/repr-backup.sh

​
Weekly Backup (Linux)

Same script, same cron job. Works everywhere.

​
Weekly Backup (Windows)

Use Task Scheduler to run: 
repr data backup --output %USERPROFILE%\Dropbox\repr-backup-%date:~-4,4%-%date:~-10,2%-%date:~-7,2%.json

​
Restore from Backup

You got a new laptop. Or your hard drive died. Or you accidentally deleted your stories. Time to restore from backup.

​
Full Restore (Replace Everything)

# This will replace all your current repr data
repr data restore ~/backups/repr-backup-2026-01-05.json --replace
Output: 
Restoring from backup...

⚠ This will replace all current data
⚠ Current stories: 12 files will be deleted
⚠ Backup contains: 45 files

Continue? [y/N] y

✓ Restored 45 stories
✓ Restored 3 profiles
✓ Restored config
✓ Restored queue (8 pending commits)

Restore complete.
Everything is back.

​
Merge Restore (Keep Both)

Already have some stories on the new machine? Want to merge with your backup? 
repr data restore ~/backups/repr-backup-2026-01-05.json --merge
Output: 
Restoring from backup (merge mode)...

Current stories: 12
Backup stories: 45

Merging:
  • 33 new stories will be added
  • 12 existing stories will be kept
  • 0 conflicts (no duplicates found)

✓ Merged 45 stories from backup
✓ Total stories: 45 (12 existing + 33 new)

Merge complete.
Now you have stories from both the backup and your local machine.

​
Migrating to a New Computer

Full workflow for setting up repr on a new machine with all your existing data.
1

Install repr on new machine

brew install repr
# Or: pipx install repr-cli
2

Copy your backup to the new machine

Transfer your backup file (via USB, cloud storage, etc):
# Example: download from Dropbox
cp ~/Dropbox/repr-backup.json ~/Downloads/
3

Restore from backup

repr data restore ~/Downloads/repr-backup.json --replace
Your stories, config, and settings are now on the new machine.
4

Initialize with your new repos

Point repr at your code folder on the new machine:
repr init ~/code
This scans for repos. The repos don’t have to match your old machine—repr works with whatever you have locally.
5

(Optional) Re-authenticate

If you use cloud sync:
repr login
This connects the new machine to your account. Your auth token is not in the backup (it’s in your OS keychain), so you’ll need to log in again.
6

Verify everything works

repr stories
repr doctor
Check that all your stories are there and repr is healthy.

​
Incremental Backups

Don’t want to back up everything every time? Use incremental backups: 
# Backup only stories created since last backup
repr data backup --output backup-incremental.json --since "1 week ago"
This creates a smaller backup with just recent work. Combine with periodic full backups:
  • Weekly: Incremental backup (fast, small)
  • Monthly: Full backup (slower, complete)

​
Checking What You Have

Want to see how much data you’re storing before backing up? 
repr data
Output: 
Local Data Storage

Stories:
  Count: 45 files
  Size: 2.1 MB
  Location: ~/.repr/stories

Profiles:
  Count: 3 files
  Size: 340 KB  
  Location: ~/.repr/profiles

Cache:
  Size: 120 KB
  Location: ~/.repr/cache

Config:
  Size: 8 KB
  Location: ~/.repr/config.json

Queue:
  Pending commits: 23
  Size: 45 KB
  Location: ~/.repr/queue

Total: 2.6 MB
Not much! Your stories are text files, so they don’t take up much space.

​
Housekeeping

​
Clear Cache (Keeps Stories)

Repr caches some data for performance. You can safely clear it: 
repr data clear-cache
Your stories are untouched. Only temp files are removed.

​
Export Individual Stories

Want to back up just one important story? 
# View the story
repr story view 01ARYZ6S41TSV4RRFFQ69G5FAV

# Export as markdown
repr story view 01ARYZ6S41TSV4RRFFQ69G5FAV --format md > important-story.md

# Or copy the raw JSON
cp ~/.repr/stories/01ARYZ6S41TSV4RRFFQ69G5FAV.json ~/backup/

​
Backup Best Practices

​
1. Backup Before Major Changes

Before doing anything destructive: 
repr data backup --output ~/temp-backup.json
Then make your changes. If something breaks, restore from ~/temp-backup.json.

​
2. Keep Multiple Backups

Don’t overwrite the same backup file. Use dated filenames: 
repr data backup --output ~/backups/repr-$(date +%Y-%m-%d).json
This way you have history. If you realize you need something from 2 weeks ago, you have it.

​
3. Test Your Backups

Occasionally verify your backup actually works: 
# Create a test directory
mkdir ~/test-restore
cd ~/test-restore

# Restore to a temp location
repr data restore ~/backups/repr-backup.json --output ./test

# Verify stories are there
ls -la ./test/stories/
A backup you haven’t tested is not a backup.

​
4. Use Cloud Sync as Secondary Backup

If you use repr’s cloud sync, your stories are already backed up to repr.dev. But don’t rely on this alone—also keep local backups. Cloud sync is great for:
  • ✅ Multi-device access
  • ✅ Automatic synchronization
  • ✅ Secondary backup
But local backups are better for:
  • ✅ Complete control
  • ✅ Offline access
  • ✅ No dependency on repr.dev being online
Use both.

​
Disaster Recovery

Your hard drive crashed. Everything is gone. What do you do?

​
If You Have Cloud Sync

# Install repr on new machine
brew install repr

# Login with your account
repr login

# Pull all stories from cloud
repr sync
You’re back up and running. This is why cloud sync is valuable even if you primarily work offline.

​
If You Only Have Local Backups

# Install repr
brew install repr

# Restore from your most recent backup
repr data restore ~/Dropbox/repr-backup-2026-01-05.json --replace
All your stories are back.

​
If You Have Neither

This is bad. Your stories are gone. But your git history still exists! You can regenerate: 
# Initialize repr
repr init ~/code

# Regenerate stories from scratch
repr generate --since "6 months ago" --local
You’ll get new stories based on your git history. They won’t be exactly the same as before (LLMs are non-deterministic), but you’ll recover most of your work. This is why backups matter.

​
Migration Checklist

Moving to a new machine? Follow this checklist:
  • Create fresh backup on old machine: repr data backup
  • Transfer backup file to new machine (USB, cloud, etc)
  • Install repr on new machine
  • Restore from backup: repr data restore --replace
  • Initialize with new repos: repr init ~/code
  • (Optional) Re-authenticate: repr login
  • Verify: repr stories and repr doctor
  • Test generate: repr generate --local
  • Set up new automated backup schedule

​
The Bottom Line

Repr stores everything as human-readable JSON files in ~/.repr/. You can back them up like any other files. But using repr data backup is easier because:
  1. It bundles everything into one file
  2. It’s guaranteed to capture the right structure
  3. It’s designed to work with repr data restore
Back up regularly. You’ll thank yourself later.