晋升评审没过，最常见的原因不是做得少，是材料里全是「我做了什么」而没有「这带来了什么改变」。而这一点自己看不出来——因为都是自己写的，每一条你都知道背后有多少工作量。 工具做的判定很机械，也正因如此有用：把你列的每一条事迹逐条查三样——有没有数字、有没有影响范围的表述、有没有说清是你做的。三样都有才是 🟢。为什么是这三样：评审本质上是比较，而评委并不了解你的日常；没有数字无法比较，没有范围看不出职级，归因不清不知道该算谁的。 再按六个常见评分维度看哪几维是空的——空着通常不是你没做，是写材料时没往那个方向想，逐条自问一遍往往能想起两三件本来做过却没写的事。 最后是评委最常问的五个追问，它们的共同点是：问的都不是你做了什么，而是你怎么想的。材料决定你能不能进答辩，回答决定你能不能过。$2 解锁逐条改写、按目标职级的材料结构、以及模拟追问与答法。Promotion packets usually fail not because you did too little, but because they describe what you did rather than what changed. You can't see this in your own writing — you know how much work sat behind every line. The check here is mechanical, which is exactly why it helps: each claim is tested for three things — a number, a statement of reach, and clear attribution. All three, and it's green. Why those three: a review is fundamentally a comparison, and the panel doesn't know your day-to-day. Without a number there's nothing to compare, without reach there's no signal of level, and without attribution nobody knows whose work it was. Then six common assessment dimensions are checked for gaps — a blank one usually means you didn't think in that direction while writing, and going through them tends to surface two or three things you actually did but left out. Finally, the five questions panels ask, which have one thing in common: none of them ask what you did, they ask how you thought. The packet decides whether you get to defend; the answers decide whether you pass. $2 unlocks line-by-line rewrites, a packet structure for your target level, and simulated follow-ups.
逐条改写（把「做了什么」改成「改变了什么」，改写前后对照）+ 按目标职级的材料结构（哪几条放前面、每条给多少篇幅）+ 十道模拟追问与答法框架 + 答辩现场的时间分配与被打断时的处理。Line-by-line rewrites turning activity into change with before/after pairs, a packet structure for the target level, ten simulated follow-up questions with answer frameworks, and how to manage time and interruptions in the room.
本工具做材料结构的检查与答辩准备，不代表任何公司的实际评审标准——各家的职级定义、评分维度与晋升配额差别很大，请以你公司的职级说明与主管的口径为准。逐条判定基于关键词匹配（是否出现数字、范围表述、归因用词），会有误判，请自行复核。晋升结果还受配额、时机、团队情况等你无法控制的因素影响，材料做好不等于一定通过。请勿在材料中夸大或虚构成果——评审通常包含交叉验证，而被发现的代价远大于一次没过。你填写的内容不入库。This checks packet structure and prepares you to defend it. It does not represent any company's actual criteria — level definitions, assessment dimensions and quotas vary widely; rely on your own company's ladder and your manager. The per-claim check is keyword-based (presence of numbers, reach language and attribution words) and will misfire; verify it yourself. Outcomes also depend on quota, timing and team context beyond your control, so a strong packet is not a guarantee. Never overstate or fabricate results — reviews typically include cross-checks, and being caught costs far more than one failed cycle. Nothing you enter is stored.