Fable
Three mirrors revealed what was real and what was not in the answers
There was a Q&A teahouse in the city, where guests came with ledgers, prescriptions, and contracts to consult the proprietor. At first the proprietor answered from experience, the guest nodded, and the teahouse counted it as a completed order.
As the business grew, apprentices began helping by checking files before answering. The answers became smoother and smoother, but smoothness does not equal reliability; some replies were unrelated to the question, and some sounded reasonable but had no evidence.
The shopkeeper first required apprentices to write a few more sentences of explanation when answering. Once the explanations grew longer, it became harder for customers to spot errors, and apprentices became better at making unchecked content sound as if it had been checked.
Once a merchant signed a bad contract on the teahouse's advice. In the retrospective, everyone argued only about whether the answer looked like something an expert would write, but no one could separately judge whether it answered the question, whether it was based on the materials, or whether it smuggled in dangerous advice.
An old tea master brought out three mirrors. The first checked whether the question and answer were relevant; the second checked whether the answer was supported by the case files; the third checked whether the materials found were actually useful.
He had each apprentice look in the mirror after answering. The mirror wouldn't answer for the apprentice, but it would leave scores, reasons and failure locations, letting the shopkeeper know whether to fix the grading, the questioning method, or the final expression.
Later, the teahouse also turned toxic advice, off-topic questions, and unsupported assertions into checklist items. Different businesses used different combinations of mirrors—you can't let one ruler pretend to fit every scenario.
Before new apprentices entered the hall, they first had to answer old customers' questions and pass the mirror inspection; after launch, customer complaint samples continued to flow back and became the next round of tea evaluation material. Only then did the shopkeeper understand that complex Q&A cannot just ask whether the answer looks beautiful; it must also reflect where it came from, whether it hit the point, and whether it might harm someone. The old tea master also required that mirror inspection results feed back into the process. If relevance was poor, change how questions are routed; if case-file support was weak, change how files are searched; if there was dangerous advice, add interception and review. Looking in the mirror was not for assigning scores, but for knowing where to make the next cut. The teahouse later hung the mirrors behind the counter, where apprentices could see them every day. They reminded everyone that smoothness is only surface; only answers that can be held up to questions, materials, and safety boundaries are worth giving to customers. Later, during the review, everyone wrote this lesson into daily rules: first look at where the scene actually changed the outcome, then decide what to keep, what to fix, and what to restrict.