结构化面试真正的难点不是不会说，是时间分配。三题二十分钟，想两分钟答三分钟，第三题就只剩喘气的时间——而考官对最后一题的印象往往最深。
所以这个工具先算时间：按题量与总时长算出每题思考多少秒、作答多少秒、什么时候必须收尾（思考约三成、作答约七成、留 20 秒缓冲）。练的时候必须掐表——不掐表的练习几乎没有意义，因为人会系统性地高估自己的语速。
然后按题型给框架。贴一道例题进来，代码识别它属于哪一类——综合分析（怎么看）、组织管理（怎么做一件事）、应急应变（突发怎么办）、人际关系（和人怎么处）、自我认知（你是谁）——给出对应的答题结构和这一类最常见的失分点：
· 综合分析最常见的失分是只说好话，把对策说成正确但空洞的话；
· 组织管理的区分度全在细节，「提前三天在哪个渠道通知、谁负责签到」胜过十句「精心组织周密安排」；
· 应急应变最常见的失分是先解释后处理——现场还没稳定就开始讲道理，顺序错了后面说得再好也不成立；
· 人际关系几乎从不考谁对谁错，一旦开始论证自己没错就已经偏了；
· 自我认知里，「奉献」「情怀」说多了会变成扣分项。
框架的作用是让你在紧张时不至于空白，不是让你套模板——考官一天听几十个人，套话听得出来。
各地各系统的题型分布、时间设置与评分方式差别很大且逐年调整，这里的时间分配是通用估算，请以你报考岗位的实际公告为准。$3 解锁五类完整框架、按考区侧重的答题重点、计时模拟与自评表、以及答非所问的诊断方法。
The hard part of a structured civil-service interview isn't finding words — it's allocating time. Three questions in twenty minutes: plan two and speak for three, and the last question gets what's left of your breath — and the last answer is often the one remembered.
So this starts with the clock: from the question count and total time it computes seconds to think, seconds to speak, and when to start closing (roughly thirty percent thinking, seventy speaking, twenty seconds of buffer). Practise with a timer — untimed practice is close to useless, because everyone overestimates how fast they speak.
Then the structures. Paste a sample question and the code identifies its type — analysis (what do you make of it), organisation (how would you run this), contingency (something has gone wrong), interpersonal (a colleague, a supervisor, the public), self-assessment (who are you) — and gives the matching structure plus the way that type is usually lost:
· Analysis answers fail by being uniformly positive and empty of anything actually thought;
· Organisation answers are separated entirely by detail — who notifies whom, through which channel, three days ahead;
· Contingency answers fail by explaining before stabilising — get the order wrong and nothing after it counts;
· Interpersonal questions almost never ask who was right, and arguing that you were is already off track;
· Self-assessment answers sink under abstract virtue words.
A structure exists so you don't go blank under pressure — not so you can recite a template. Panels hear dozens of people a day; recitation is audible.
Formats, timings and scoring differ by region and change yearly, so the timing here is a general estimate — follow your own announcement. $3 unlocks all five structures, regional emphases, a timed mock with a self-scoring sheet, and how to diagnose answering the wrong question.
五类题型的完整框架与各自的失分点、按考区与岗位侧重的答题重点、计时模拟流程与自评表（录像后逐项打分）、以及「答非所问」的诊断方法——这是最常见也最容易被忽略的失分原因。All five question types with their structures and failure modes, what different regions and posts emphasise, a timed mock protocol with a self-scoring sheet, and how to diagnose answering the wrong question — the most common and least noticed failure.
本工具提供的是通用的答题结构与时间估算，不代表任何考区、任何岗位的实际考试形式或评分标准。 各地、各系统的题型分布、题量、时长、是否听题、是否有草稿纸、评分维度与权重差别很大且逐年调整——请以你报考岗位的招录公告、面试通知与官方说明为准。时间分配（思考约三成／作答约七成／20 秒缓冲）是本工具采用的通用估算，不是规定。本工具不提供任何真题、内部信息或「押题」，也不承诺任何成绩或结果。框架用于组织你自己的内容——照搬模板与套话在实际评分中通常是减分项。你填写的内容不入库。
This offers general answer structures and a timing estimate; it represents no region's or post's actual format or scoring. Question types, counts, durations, whether questions are read aloud, whether notes are allowed, and the scoring dimensions all vary considerably and change from year to year — follow the recruitment announcement and interview notice for the post you applied to. The timing split (about thirty percent thinking, seventy speaking, twenty seconds of buffer) is a convention used here, not a rule. This tool provides no exam questions, no insider information and no predictions, and promises no result. The structures exist to organise your own material — reciting templates typically scores worse, not better. Nothing you enter is stored.