Affected-Customer Identifier
When an incident hits, build the exact list of impacted customers to notify — so comms reach everyone affected and nobody who isn't — with a hard approval gate before anyone gets emailed.
Enter an incident scope, get a filtered list of impacted customers with live counts, have the incident manager approve it, then lock it and hand a clean export to your broadcaster — with the exact criteria logged for the post-incident report.
Before you start
- A customer dataset (CSV or Google Sheet) with region, product, plan/version, and contact email
- A free Supabase account
- A free Vercel account
- A free Resend account (only if you want the tool to email the locked list)
The problem this kills
When an outage or incident hits, the worst thing you can do is guess who to notify. Notify too few and affected customers find out from Twitter before they hear from you. Notify too many and you spook customers who were never touched, flood your support queue, and erode trust in every future alert you send.
Today that list usually gets built by hand: someone filters a spreadsheet under pressure, eyeballs which regions and plan versions are in scope, copies emails into a blast, and prays. There's no record of the criteria used, no second pair of eyes before send, and no way to prove later who you contacted and why.
This tool replaces the panicked spreadsheet filter with a calm, repeatable process — and puts a mandatory human approval gate between "here's the list" and "send."
What you'll build
A small internal web app where your team:
- Loads your customer dataset (region, product, plan/version, contact email) from a CSV or Google Sheet.
- Enters an incident scope — the affected region(s), product(s), plan(s)/version(s), or a pasted list of affected account IDs.
- Sees the dataset instantly filtered to only impacted customers, with live counts (total impacted, by region, by plan) and a preview table.
- Lets the incident manager review and approve that exact list before it can be used.
- Locks the approved list, hands it to your broadcaster (or exports a clean CSV), and logs the precise scope and criteria for your post-incident report.
What's inside the Implementation Plan
The plan is a complete, paste-and-go runbook for an AI coding agent (Claude Code). You don't write code — you paste, answer questions, and approve.
It opens by interviewing you about your business — your incident process, the systems and spreadsheets you use, the real field names and ID conventions in your customer data, your typical and peak incident volumes, and your exact approval rules — so the tool is tailored to how you actually run comms, not a generic template. The agent reflects a short spec back to you and waits for your thumbs-up before it builds anything.
From there it walks step by step through standing up the database, the import, the scope-and-filter screen, the live counts, the approval gate, the lock-and-handoff, the audit log, and the CSV export — each step ending in a ready-to-paste prompt.
The governance it includes (this is the point)
- Login so only your team can open the tool.
- Row-level security so each organization only ever sees its own customers and incidents.
- A complete audit trail — who entered the scope, who approved the list, the exact filter used, and when.
- A hard human-in-the-loop approval gate — the tool drafts the affected list; the incident manager reviews and approves; only then can it be locked and used for any outbound communication.
- Duplicate guards — the dedupe key is customer ID within an incident, so the same person can't land on the list twice and can't be processed twice.
Who it's for
Incident managers, support leads, and operations people who run outage and incident communications and need to be confident the right customers — and only the right customers — get notified.
You've got this. Open the plan, paste the first prompt, and let the agent interview you.