Reopen & Repeat-Contact Analyzer

The problem this kills

Your team closes tickets all day, but the same problems keep walking back through the door. A customer gets a "resolved" reply, then writes again two days later about the exact same thing. A ticket gets reopened, then reopened again. Each of those is invisible work - effort spent re-doing something that should have been fixed once.

Most support tools count volume (how many tickets) and speed (how fast you closed them). Almost none tell you the thing that actually drains your team: repeat work. Which issues keep coming back? Which fixes don't stick? Which root causes are quietly generating a slow drip of reopens and "me too, still broken" replies?

Without that picture, you keep treating symptoms. You coach agents on closing faster when the real problem is a broken refund flow, a confusing setup email, or a policy nobody understands. You're mopping the floor with the tap still running.

What you'll build

A small, private web app for your support-ops team that:

• Loads your ticket history from a CSV export or Google Sheet (no integration required to start).
• Detects reopened tickets - any ticket that was marked resolved and then came back to life.
• Detects repeat contacts - the same customer reaching out again about the same issue inside a window you define (so a genuinely new problem doesn't get miscounted).
• Links each repeat back to its original ticket so you can see the full chain, not just isolated events.
• Clusters the reasons into plain-language buckets ("refund didn't process," "setup instructions unclear," "shipping ETA wrong") so patterns jump out.
• Puts a human in charge: you, the analyst, review the clusters, fix any mislabels, and approve the findings before they count.
• Exports and emails a clean "top repeat-work drivers" report - ranked by how much repeat work each driver is causing - ready to hand to a manager for a coaching or process-fix decision.

What's inside the Implementation Plan

• It starts by interviewing you about your business. Before a single line is built, the plan makes the AI agent ask about your current process, your helpdesk and exports, your real field names, your volumes, and your own definition of a "repeat contact." It reflects a short tailored spec back to you, you give a thumbs-up, and only then does it build. This is the opposite of a generic template - the tool is shaped to your data.
• A step-by-step build you paste into Claude Code, one prompt at a time.
• A data model tuned to how reopens and repeat contacts actually look in your export.
• Reopen + repeat-contact detection logic that respects your "same issue vs. new issue" window.
• Reason clustering with a human review-and-approve screen.
• The "top repeat-work drivers" report, with on-screen view, CSV export, and email delivery via Resend.
• A no-integration fallback so you can ship this entire tool this weekend using only an export file.

The governance it includes (this is the point)

This isn't a throwaway script - it's an internal tool you can trust and defend:

• Login so only your team can open it.
• Row-level security so each organization only ever sees its own data.
• A full audit trail - who loaded what, who approved which findings, and when.
• A human-in-the-loop approval gate - the AI drafts the clusters and the report, but nothing becomes an official "driver" until you review and approve it. No automated finding ever feeds a coaching decision unreviewed.
• Duplicate guards so loading the same export twice can't double-count the same reopen or repeat.

Who it's for

Support-ops analysts and support managers who are tired of closing the same ticket twice and want hard evidence of where the repeat work comes from - so they can fix the root cause, justify a process change, and coach with data instead of hunches. No coding background needed.

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