N NoBullshit.wiki 中文

Breakfast shop equipment technician - the three-stage repair call scripts: report symptoms on the phone, settle price before work starts, and test under load when it is done, in read-aloud form

Works with:Claude / GPT-4 classCategories:Breakfast ShopLevel:Beginner Updated

You are the equipment lead at a breakfast shop. The steamer, dough
mixer, fryer, or soy grinder has failed and I need to call the
repair technician. Give me three read-aloud scripts plus a
"how to describe the symptom" table for each machine.

Stage one: the repair call (symptoms, not diagnosis)
- Criteria: symptom + when it started + what I already ruled out.
  Never "I think X is broken" - a self-diagnosis skews the parts
  he brings; he shows up with one part, finds a different cause,
  and the wasted trip costs a second call-out fee
- Read-aloud opener: "Mr. Chen, the shop steamer ran an hour this
  morning without steaming; it was fine yesterday. Water level and
  power I already checked. What time can you come?"
- Symptom table per machine (I fill the symptom, you complete the
  sentence):
  Steamer: no steam / steaming slower than usual / water level
  will not rise
  Dough mixer: abnormal noise / belt slipping, drum not turning /
  dead stop
  Fryer: oil will not reach set temperature / thermostat jumping /
  breaker tripping
  Soy grinder: gritty output / abnormal noise / jammed and stopped

Stage two: settle price on arrival (three things before any work)
- "Can it be repaired, or does it need a part"
- "If a part: what does it cost, what spec, and is it the same
  spec as what is on my machine"
- "How is labor charged - and if I need it by tomorrow morning,
  what is the rush surcharge"
- Two standing rules: parts get quoted before installation; the
  old part stays with me
- When he says "not worth repairing, replace the whole unit":
  "Which part exactly has failed? How many times has this part
  been repaired before?"

Stage three: acceptance (under load, never a no-load spin)
- Steamer: full of trays, timed from cold start to normal steam -
  anything slower than my recorded [baseline] is not fixed
- Fryer: timed from cold oil to set temperature
- Dough mixer: a full batch, listened to start to finish
- Criteria: acceptance result and warranty period go into a text
  message from him - not a voice note

Output format:
- A read-aloud version for each stage, spoken short sentences
- The symptom table: machine / symptom (I fill) / complete
  sentence for the repair call
- End with a list of forbidden lines, each as "do not say X -
  because Y"

Constraints:
- No invented brands or model numbers; no "typical repair cost"
  figures - prices are asked, never guessed
- No "build a good relationship with the technician" advice that
  is not part of a script
- I am not providing baseline times; leave [ ] blanks to fill from
  this shop's own records

Why it is written this way

All three stages serve one objective: fix it right the first time. The breakfast shop's repair window is brutally narrow - two afternoon hours, or before the next dawn opening - and a repair that does not hold costs an entire morning service. That is why the call reports symptoms instead of a diagnosis: if you are certain it is the heating element, the technician brings only a heating element, arrives, finds the water-level sensor, and the wasted trip comes out of your wallet in call-out fees and half a day. "What I already ruled out" matters just as much - it burns off every check he would otherwise walk you through by phone, so he can start work the moment he arrives.

Price talk sits before the work because after the work there is no negotiating position left: the machine is bolted back together, the morning rush is waiting for power, and you pay whatever the invoice says. The three questions are ordered by information value: repairable-or-part decides this visit's cost; the part's spec decides whether the quoted price is sane (spec differences swing part prices hard); the rush surcharge decides whether tonight is worth it. Keeping the old part is the cheapest evidence mechanism that exists - whether a part was replaced at all, and whether it was the part that was quoted, is settled by the lump of metal in your hand.

Load-based acceptance comes from scar tissue: a steamer that passes a no-load test can still fail with full trays - heating element degradation only shows under load. The baseline must be this shop's own recorded time (which is exactly what the daily three-reading habit buys you); any "how long is normal elsewhere" is meaningless, since equipment age, water hardness, and the shop's voltage all differ. The warranty goes into a text message rather than a verbal promise or a voice note because text is searchable - when the same fault returns three months later, scrolling the chat beats trusting anyone's memory. And the constraint "prices are asked, never guessed" exists specifically to wall off the model's favorite sin: quoting an "average" repair price. Repair costs swing enormously by region and machine type, and every "usually around..." is fabricated.

Variables

Variable Notes
Acceptance baselines The under-load time standards per machine - from this shop's own startup records; use the same numbers before and after the repair

repair callsrepair acceptancespoken scripts