"Waiting on Reply" Follow-Up Tracker

The problem this kills

Shared inboxes leak threads. A customer replies, it slides three screens down, and three days later you realize nobody answered. Or you're the one waiting on the customer — and the polite nudge that would have closed the deal never went out because no one was tracking the clock. The "ball is in our court" threads and the "we're waiting on them" threads are invisible in a busy inbox, so they quietly rot until someone complains.

The usual fix is a person scrolling the whole inbox every morning, eyeballing who wrote last, and guessing what's overdue. It's slow, it's inconsistent, and the single most common mistake — misreading who actually sent the last message — sends a "just checking in!" to a customer who is, in fact, still waiting on you. Embarrassing, and avoidable.

What you'll build

A small, private web app for your team that takes your inbox threads, figures out the direction (who sent the last message) and the age (how stale it is), and splits everything into two crisp lists: "We owe a reply" and "Follow up with the customer." For each thread it drafts the right kind of nudge — an apology-and-answer for the ones you owe, a gentle check-in for the ones you're waiting on — and then it stops. Nothing sends until a human reads the draft and clicks approve. Approved nudges go out through Resend and get logged, so the same thread never gets nudged twice by accident.

What's inside the Implementation Plan

• It starts by interviewing you about your business. Before a single line is built, the plan has the AI agent ask about your actual inbox, your fields and naming, your follow-up thresholds, your escalation rules, and your messiest edge cases — then it reflects a short tailored spec back and waits for your thumbs-up. You get a tool shaped to how you work, not a generic template.
• A step-by-step build you paste into Claude Code, one ready-to-copy prompt at a time.
• The exact data model for threads, directions, ages, and nudge history — tuned to your column names.
• The logic to correctly detect who sent the last message (the bug everyone hits), compute thread age against your thresholds, and build both worklists.
• Per-thread nudge drafting with a human approval gate before anything sends.
• A follow-up cap so threads escalate or close instead of being nudged forever.
• A "No API yet?" fallback: import a CSV/Google Sheet today and export a clean follow-up worklist CSV — no inbox integration required to get value.

The governance it includes (this is the point)

This isn't a script that blasts emails. It's an internal tool built like a grown-up system:

• Login so only your team can open it.
• Row-level security so each organization only ever sees its own threads.
• A complete audit trail — who approved which nudge, and when.
• A hard human-in-the-loop approval gate: the AI drafts, a person reviews and approves, and only then does it send and write to your records.
• Duplicate guards keyed on thread id + nudge count, so the same thread can't be nudged twice.

Who it's for

Shared-inbox teams — support, success, sales ops, any group working a help@ or info@ mailbox — who keep losing threads in the scroll and want a reliable daily "what needs a reply, what needs a nudge" worklist without buying another platform.

You've got this — paste the first prompt and let the interview tailor the rest.