Workflow

Build a support-routing sandbox before connecting a channel

Turn a fast customer-support demo into a small table of fictional tickets, expected destinations and escalation conditions.

4 min estimated readStart hereEvidence baseline

THE SHORT ANSWER

Test the classification and permission boundary with fictional tickets before connecting real conversations.

Extract the task from the demo

Mckay Wrigley’s archived post describes a quick support agent that routes tickets to Slack channels. Its build-time and code-generation claims are not reliability evidence, and the video was not reviewed. Slack’s current message documentation makes channel identity and permissions explicit. That gives a concrete boundary for a sandbox task: selecting a destination is different from being allowed to send there.

Create a fictional ticket table

Write six short requests: an ordinary question, two topics in one message, an unknown product, a request containing sensitive-looking dummy data, a duplicate and a request for a restricted destination. Give each a human-written expected route or escalation decision. Include a reason and the smallest information the recipient needs. Start by asking the system to return a proposal without sending anything.

Evaluate the complete decision

Check destination, unnecessary disclosure, duplicate handling and whether uncertainty produces a human handoff. A wrong route is not fixed by a friendly acknowledgement. If the workflow is later connected under authorization, verify the actual returned channel and message record as well. This is an original sandbox protocol, not a tested support system, adoption result or instruction to contact real users.

Put it to work.

  • Use fictional tickets.
  • Label routes and escalation.
  • Keep proposal and send separate.
  • Check duplicates and disclosure.

Go to the source.

Original material, with its own scope and publication date. This note is an explanation, not a CooScope performance test.

Source discovery: @mckaywrigley

Source published · Editorial review

Archived English post text; primary sources reviewed separately. Generated annotations, media and missing replies are outside the evidence scope.

  1. @mckaywrigley — archived X post (2024-09-09)
  2. Slack — chat.postMessage referenceLink / source review: 2026-10-15
Evidence baseline and source-review dates are different. A recent link check does not make an older study a current product test. Read our method →

Keep the thread going.

Explore this topic →