I’m building a workflow where certain actions, like blocking an IP or disabling a user account, should only happen after a human reviews and approves. I want to pause the story and wait for input before proceeding. Is this something Tines supports natively, and how do people typically set it up?
This is a great use case for Tines Pages, which lets you create simple web forms or prompt interfaces that a human can interact with. The story sends a notification (via Slack, email, etc.) with a link to a Page where the reviewer can approve or reject, and the story waits for that input before continuing down the appropriate path. Nothing like blocking an IP or disabling an account fires until someone says go.
The nice thing is your workflow naturally splits into three phases:
- Automated prep (the safe stuff): enrich the alert, look up the user, pull risk scores.
- Human checkpoint: the Page presents that context and waits for a decision.
- Automated execution (the risky stuff): only runs on the branch that matches the approval.
Because the destructive action sits after the checkpoint, it structurally can’t happen without sign-off.
You can customize the Page to display all the relevant context the reviewer needs (the IP, who’s affected, reputation scores, and so on), keeping everything in one place. Add a button so they can submit their decision, and give each button a submission value (like approve or deny) that gets passed into the event. The approval and rejection paths in your story are then just separate branches, so you have full control over what happens in each case. Prefer to keep people in chat? You can run the same pattern with an approve/deny prompt in Slack or Teams instead.
A couple of tips!
- Enrich before the prompt, not after. Give the reviewer everything they need on one screen so the checkpoint is a quick decision, not a second investigation.
- Add timeout or escalation logic. A story that blocks forever on a pending approval just silently stalls. Decide what happens if no one responds: escalate to the next person, or halt safely and alert.
This pattern is well-supported and commonly used for sensitive actions in security workflows. A few doc resources to get you moving:
- Docs: List of page elements and Building apps and flows cover how to configure inputs and wire them into your story logic.
- Story Library templates you can adapt quickly, like these examples: Send a prompt to a Slack channel or user & take action, Respond to approve and deny requests via Slack, and Analyze Elastic alerts, block IPs, and notify in Slack and Tines Cases (close to your exact use case).
- Tines University has guided examples of human-in-the-loop patterns if you’d like a walkthrough.