Suppression & Unsubscribe Reconciler

The problem this kills

When somebody unsubscribes, it should be over. But in real life you mail from more than one place - your ESP (the email tool that sends your campaigns), your CRM, a support or help-desk tool that fires its own emails, and a handful of "please take me off your list" replies that someone handled by hand. A person who opts out of the newsletter in your ESP is very often still sitting active in the CRM, still getting "we miss you" notes from the support tool, still on the spreadsheet someone forgot about.

That's not just sloppy - it's a real legal liability. Under CAN-SPAM (US) and CASL (Canada), an opt-out has to be honored everywhere, fast. Mailing someone who unsubscribed is exactly the thing that gets you complaints, spam-folder placement, and fines. And the proof you need - "we suppressed this person on this date and here's who approved it" - lives nowhere, because each tool only knows about its own list.

This Implementation Plan replaces the quarterly "let me VLOOKUP four exports together and pray" ritual with a small, private internal tool. It pulls every suppression list into one place, treats any opt-out as global, finds everyone who's still mailable somewhere they shouldn't be, and won't make the master list authoritative until a human approves it.

What you'll build

A login-protected web app for your email/marketing ops team that:

• Imports a suppression/unsubscribe export from each system you mail from (ESP, CRM, support tool, manual list), plus your current active mailing list.
• Normalizes every email (lowercase, trim, strip the obvious junk) so the same person matches across files no matter how each tool stored them.
• Merges all opt-outs into one master suppression list, keyed on the lowercased email - one person, one suppression record, no matter how many lists they appear on.
• Flags the dangerous gaps: anyone who is suppressed in at least one system but still active in your mailing list or active in another system - your actual CAN-SPAM/CASL exposure, listed plainly.
• Routes the whole reconciliation report through a human approval gate: an admin reviews who's suppressed where versus still mailable and approves the master list before it's treated as authoritative.
• Exports the master suppression CSV plus a per-system "to-remove" CSV for each tool, in the exact columns that system expects, so you can actually go clean each one.

It's built to fit your stack, because the very first thing it does is interview you about which systems you mail from and what their exports look like.

What's inside the Implementation Plan

The plan is a single file you paste into an AI coding agent (Claude Code), and it builds the tool with you step by step. No prior coding needed.

• It starts by interviewing you about your business. Before writing a line of code, the plan has the agent ask which systems you mail from, what each suppression export actually looks like (the real column names), how big your lists are, who's allowed to approve, and the messy edge cases - role addresses, plus-addressing, hard bounces, re-subscribes. Then it reads back a short tailored spec and waits for your thumbs-up. You get a tool shaped to your setup - not a generic template.
• A clear definition of done so you know exactly what "finished" looks like.
• Every build step ends with a ready-to-copy prompt you paste into the agent.
• A "No API yet?" fallback so you can build the whole thing today using CSV/Google Sheet imports and clean CSV exports - no integration with your ESP or CRM required.
• A verification checklist so you can prove it works before anyone relies on it.

The governance it includes (this is the point)

This isn't a toy. Every runbookify plan builds in the controls that make a compliance tool safe to rely on:

• Login so only your team can open it.
• Row-level security so each organization only ever sees its own contacts and lists.
• A complete audit trail - who imported which file, who ran the reconciliation, who approved the master list, and when. This is exactly the proof a CAN-SPAM/CASL question demands.
• A hard human-in-the-loop approval gate - the tool merges and flags, but the master suppression list is not authoritative until an admin reviews and approves it. The AI drafts; a person commits.
• Duplicate guards keyed on the lowercased email, so the same person can never create two suppression records and the same file can't be processed twice.

Who it's for

Email and marketing ops people who send from more than one system and are quietly worried that an opt-out in one tool isn't being honored in the others. If you can export a CSV from each of your tools, you can run this - and you'll finally have one authoritative suppression list and a defensible record of how it was built.

You've got this - paste the first prompt and let the agent interview you.