What are Workflow Checkpoints

Checkpoints in Gumloop provide a simple way to create snapshots of your work and roll back when needed. Here’s how it works:
  • You’re always editing the live checkpoint directly
  • Create checkpoints whenever you want to save a snapshot
  • Any past checkpoint can be made live instantly
  • Every checkpoint is saved forever in your history
Simple analogy: Think of it like Google Docs. You’re always editing the real document, and checkpoints lets you create snapshots of your work you can return to at any time.

Understanding Live Checkpoint vs Saved Checkpoints

The checkpoint that’s currently running and that you’re editing. Triggers, interfaces, and webhooks all point to this. All your edits happen here and take effect immediately.

Quick Overview

Live CheckpointSaved Checkpoint
What is it?Your current checkpoint that runs everythingA snapshot saved in history
Can you edit it?✅ Yes - you’re always editing this❌ No - read-only snapshot
Runs automatically?✅ Yes - triggers, interfaces, webhooks use this❌ No - unless promoted to live
Create checkpointSave menu → “Save & Create New Checkpoint”Already exists in history
Make it liveAlready live”Make This Checkpoint Live” from menu

How Checkpoints Work

The Flow of Work

1

Edit Your Flow

You’re always working on the live checkpoint. Changes take effect immediately as you save.
2

Create Checkpoints

Use Save menu → Save & Create New Checkpoint when you want to save a snapshot of your current work.
3

Continue Working

Keep editing the live checkpoint. Your snapshot remains safely stored.
4

Roll Back if Needed

If something goes wrong, promote any past checkpoint to instantly make it live.

Real-World Example: Building a Slack Notification Flow

Monday: Build working Slack notification flow → LIVE
        Save & Create New Checkpoint (Checkpoint 1) → snapshot saved
        
Tuesday: Add email notifications → LIVE 
         Accidentally break Slack connection → LIVE broken
         Open Checkpoint History → Make Checkpoint 1 Live → LIVE restored instantly
         
Wednesday: Fix the issue properly → LIVE
           Add email feature again → LIVE
           Save & Create New Checkpoint (Checkpoint 2) → snapshot saved

Creating and Managing Checkpoints

Creating a New Checkpoint

Save dropdown showing Save & Create New Checkpoint option
To create a checkpoint to save a snapshot of your flow:
  1. Click the Save dropdown arrow
  2. Select “Save & Create New Checkpoint”
  3. Your current state is saved as a checkpoint
  4. Continue editing the live checkpoint
Pro Tip: Create checkpoints before making major changes, so you have a known-good state to return to if needed.

Making a Past Checkpoint Live

When you need to roll back to a stable checkpoint:
Checkpoint History panel showing Make This Checkpoint Live option
  1. Open Checkpoint History from Save menu
  2. Find the stable checkpoint you want
  3. Click menu (⋮) → “Make This Checkpoint Live”
  4. Triggers/interfaces immediately switch to this checkpoint

Restoring a Checkpoint

When you need to continue working from a past checkpoint, you can restore it. This replaces your current working draft with that past checkpoint, allowing you to build directly on top of it. Steps to Restore:
  1. Open Checkpoint History from the Save menu
  2. Find the checkpoint you want to restore
  3. Click the menu (⋮) → “Restore Checkpoint”
  4. Your current draft is replaced with that checkpoint, and you are now editing from it directly
Restoring is useful when you want to take a past version and continue evolving it, rather than just viewing or duplicating it.

Difference Between Restore and Make Live

ActionWhat It Does
Restore CheckpointSets that past checkpoint as your current working draft so you can continue editing it.
Make This Checkpoint LiveSwitches triggers, interfaces, and webhooks to run that checkpoint immediately, without affecting your current draft.
Example Workflow:
Checkpoint 3: Stable version
Checkpoint 4: New feature added but buggy

You decide to scrap Checkpoint 4 and keep working from Checkpoint 3:
→ Restore Checkpoint 3 → Your editor now shows Checkpoint 3 and you continue building from there
→ When ready, Make This Checkpoint Live → Production switches to the updated version
This makes Restore perfect for cases where you want to roll back your actual work in progress and build forward from a previous stable point.

Working with Triggers

Important: Triggers always run from whichever checkpoint is currently live. When you switch checkpoints, all triggers immediately use the new live checkpoint.

How Triggers Behave

  • While editing: Your saved changes affect triggers immediately (you’re editing live)
  • When creating a checkpoint: Triggers continue running, just creates a snapshot
  • When making past checkpoint live: All triggers instantly switch to that checkpoint
ActionWhat Happens to TriggersImportant Note
Edit the flow❌ No effect until savedChanges are local only
Save the flow✅ Changes apply immediatelyTriggers use saved changes
Save & Create New Checkpoint✅ No change - just saves snapshotTriggers keep running
Make past checkpoint live✅ Triggers switch to that checkpointInstant switch
Add new trigger + Save✅ Active immediatelyMust save to activate
Delete trigger + Save✅ Removed immediatelyMust save to deactivate

Working with Interfaces

Interfaces work exactly like triggers - they always serve the current live checkpoint:
  • Users always see the live checkpoint
  • Changes to interfaces take effect immediately
  • When you promote a past checkpoint, interfaces instantly switch
  • No publish step required

Past Checkpoints Panel

Viewing Past Checkpoints

Click “View Past Checkpoints” from the Save menu to see all your checkpoints:
Checkpoint History panel
Each checkpoint shows:
  • Checkpoint number
  • Creation date and time
  • Author information
  • Live badge (if currently live)
  • Menu for actions

Available Actions for Past Checkpoints

From the menu (⋮) next to any past checkpoint:
  • Make This Checkpoint Live: Instantly switch to this checkpoint
  • Restore Checkpoint: Create a duplicate of this checkpoint as a new snapshot
  • Edit Details: Update the name or description
Checkpoint menu options

Advanced: When a Past Checkpoint is Live

This section covers an advanced scenario used by power users who need to maintain stable production while fixing issues.

What Happens When You Promote a Past Checkpoint

When you make a past checkpoint live:
  1. Triggers, interfaces, and webhooks immediately point to that past checkpoint
  2. Your current work is NOT lost - it becomes a checkpoint
  3. The past checkpoint remains read-only (you can’t edit it)
  4. You continue editing from where you left off

Returning to Your Latest Work

After promoting a past checkpoint, you might want to switch back to your recent edits:
Save menu showing Make This Checkpoint Live when past checkpoint is active
When a past checkpoint is live, the Save menu shows “Make This Checkpoint Live” - click this to:
  • Return your recent edits to live status
  • Switch triggers/interfaces back to your latest work

Example Scenario: Production Hotfix

1. Production issue occurs
2. Make stable Checkpoint 5 live → Users get working checkpoint
3. Continue fixing in your current checkpoint
4. Test thoroughly
5. Save menu → "Make This Checkpoint Live" → Deploy fix
This setup allows you to:
  • Instantly restore service with a past checkpoint
  • Fix issues without time pressure
  • Deploy when ready

Best Practices

1

Checkpoint Before Major Changes

Create a checkpoint before adding new features or making significant edits
2

Name Your Checkpoints

Add descriptions to checkpoints so you know what state they represent
3

Test After Switching

When promoting a past checkpoint, verify triggers and interfaces work as expected

Common Workflows

Quick Rollback

  1. Something breaks in production
  2. Open Checkpoint History
  3. Find last known good checkpoint
  4. Click “Make This Checkpoint Live”
  5. Production is fixed instantly

Experimenting Safely

  1. Save & Create New Checkpoint (save current state)
  2. Make experimental changes
  3. If experiment fails: promote the checkpoint back to live
  4. If experiment succeeds: create another checkpoint to save it

Frequently Asked Questions

Summary

The mental model is simple - just like Google Docs:
  • You’re always editing the live checkpoint
  • Save & Create New Checkpoint = save a snapshot (like starring a revision in Google Docs)
  • Make This Checkpoint Live = instantly switch to that snapshot
  • Triggers, interfaces, and webhooks always use whatever’s live