Q

QMy personal support ticket queue, exactly how I want it.

Q files every inbound ticket and suggests a reply. I review, edit and send. The best staff replies become canned replies that remember where they came from, and whoever has waited longest stays on top.

127.0.0.1 / inbox
The Inbox: open tickets where the requester spoke last, each row coloured by its rollupThe Inbox: open tickets where the requester spoke last, each row coloured by its rollup
Account recoveryAccount verificationProfile changesTakedowns & moderationContent correctionsAPISite bugsHow-toSite feedbackMisdirectedSpam / noiseSecurity & dataTwelve rollups, one color each, the same on every screen. Demo data: a fictional helpdesk, The Galactic Helpdesk.

Workflow

Inbound tickets are filed with a suggested reply automatically. Humans review the proposed draft and reply.

  1. Import

    Tickets, threads and requesters come in from the helpdesk export. Re-importing replaces the machine data and never touches a review.

  2. Classify

    A model files every ticket with a category, an outcome, what the requester asked for and which message solved it.

  3. Review

    Accept or correct each label with one key. The machine's answer stays next to yours.

  4. Reuse

    Good staff replies become canned replies, linked to the real tickets they came from.

  5. Inbox

    Open tickets where the requester spoke last, oldest first. Repeat requesters show their history inline.

  6. Draft and send

    Draft in Q, fill the placeholders, paste the reply into the helpdesk and mark it sent.

  7. Refresh

    New tickets come in from the helpdesk API and stay unfiled until reviewed.

Refresh feeds the next import and the workflow starts again.

Six screens, one queue

Queue and Inbox

Know who's waiting, work the oldest first, clear the queue in one sitting.

127.0.0.1 / inbox
The Inbox: open tickets where the requester spoke last, each row coloured by its rollupThe Inbox: open tickets where the requester spoke last, each row coloured by its rollup

Ticket

Read the full history, start from a draft with the details filled in, reply without hunting for links.

127.0.0.1 / tickets / 31545
A ticket page with the helpdesk thread, a Q draft with placeholders, and the classification panelA ticket page with the helpdesk thread, a Q draft with placeholders, and the classification panel

Canned replies

Analyze past responses, review three proposed drafts, craft the new standard for future replies.

127.0.0.1 / canned / suggestions
Suggested canned replies grouped by category, with the staff messages they came fromSuggested canned replies grouped by category, with the staff messages they came from

Categories and rollups

Group machine categories into rollups you can read at a glance, see what people actually ask for, decide what to fix first.

127.0.0.1 / categories
Categories screen: machine categories grouped under coloured rollup labelsCategories screen: machine categories grouped under coloured rollup labels

Requesters

Recognize repeat requesters across accounts, see their whole history, answer with context.

127.0.0.1 / requesters / 7006
A repeat requester's timeline across three accountsA repeat requester's timeline across three accounts

Dashboard

Track review progress, see which issues get resolved or refused, compare what humans and agents moved this week.

127.0.0.1 / dashboard
Dashboard: review progress, rollups by outcome, recent activityDashboard: review progress, rollups by outcome, recent activity

Around the edges

Agent access over MCP

Hand the reading and first drafts to an agent, keep a human on every reply, see exactly which work was the agent's.

Refresh from the helpdesk API

Pull in new tickets with one button, keep the queue current, never post anything back to the helpdesk.

Honest notes

Built for one queue

I built Q for my own support backlog and long evening sessions with the helpdesk open in the next tab. It fits that job closely and makes no promises about yours.

Local only

Q runs on my machine at 127.0.0.1. Ticket data stays in a local database. The only network call is the read-only refresh I start.

Not for sale

No download, no signup, no pricing. This page shows how it works. The source is available on request.