Checkpointing

Checkpointing lets you save a snapshot of your current session — conversation history, file changes, and configuration — so you can experiment freely and roll back if things don't work out. Think of it like a save point in a game.

When to Use Checkpoints

• Before attempting a risky refactor or architectural change
• When you want to try two different approaches and compare
• To preserve a working state before the AI makes further changes

Commands

• /checkpoint create [name] — Save a checkpoint (auto-generates a timestamp name if omitted)
• /checkpoint list — List all checkpoints with creation time, message count, and files changed
• /checkpoint load [name] — Restore files from a checkpoint (interactive selector if no name given)
• /checkpoint delete <name> — Permanently delete a checkpoint

What Gets Saved

• Complete conversation history
• Modified files with their content (detected via git)
• Active provider and model configuration
• Timestamp and metadata

Example Workflow

# Save current state before trying something new
/checkpoint create before-refactor

# Ask the AI to try an approach...
# If it doesn't work out:
/checkpoint load before-refactor

# If it went well, save the new state:
/checkpoint create after-refactor

# Compare what you have:
/checkpoint list

When loading a checkpoint that would overwrite current work, Nanocoder prompts you to create a backup first.

Storage

Checkpoints are stored in .nanocoder/checkpoints/ in your project directory. Each project has its own checkpoints. Consider adding .nanocoder/checkpoints to your .gitignore.

Note: Loading a checkpoint restores files immediately, but restoring conversation history requires restarting Nanocoder.