VERISTORY
交付物Deliverable

交付什么What you get

一份体检报告,中英双语 PDF,带水印和编号:One audit report, in Chinese and English, watermarked and numbered:

  1. 学生基本情况,以及这份材料呈现的是一条什么路线、什么人设Student profile and a neutral background summary
  2. 一页总览:一共几条、分别有多要紧、几个奖已经核实了At a glance: total findings, high/medium/low distribution, share of honors confirmed
  3. 每个奖项逐条标明:已在主办方名单核实,还是查不到Honor verification table — confirmed against the organizer’s list, or not verifiable against public records
  4. 主要发现:每条都标出处、给一个 0–100 的关注度分,并列出该准备什么Findings — each with a scrutiny-likelihood score out of 100, the source passage, and the evidence to prepare
  5. 追问演练:十几道细节题,真做过的孩子答得上来,没做过的会卡住Mock interrogation drill — a dozen-plus detail questions only someone who was actually there can answer
  6. 最后一页总结,外加一张可以打印出来逐条打勾的证据清单Key issues summary, plus a printable evidence checklist
  7. 这份报告怎么用,三步说清How to use the report: a three-step plan
样例一 · 材料干净
样例一 · 材料干净Sample 1 · Clean
物理方向。无刻意问题——一份"没查出大毛病"的报告长什么样。Physics track. Nothing planted — this is what a report with little to flag looks like.
样例二 · 若干需澄清
样例二 · 若干需澄清Sample 2 · Some clarifications
化学方向。文书声称的实验在活动表里找不到;重点科研拿不出证明链。Chemistry track. An experiment claimed in the essay is absent from the activity list; the flagship research has no evidence chain.
样例三 · 问题较集中
样例三 · 问题较集中Sample 3 · Concentrated
计算机方向。暑期两活动同期 83 小时/周;同一奖项级别前后矛盾。CS track. Two summer activities overlapping at 83 hrs/week; the same award described at two different levels.

这三份样例里的学生都不存在。姓名、学校、项目全是虚构的;竞赛与奖项的分布参照真实申请材料的统计结果。完整 PDF 见随信附件。No student in these samples exists. Names, schools and projects are invented; the mix of competitions and awards follows the statistical shape of real applications. Full PDFs are sent as attachments.

怎么做出来的Behind the report

怎么做出来的How it is produced

先由 AI 把材料整体过一遍:时间对不对得上、课程能不能撑起你说的科研、每段经历有没有东西可以证明、奖项在不在公开名单里、文件本身的信息有没有异常、Common App 的硬性限制有没有超。然后每一条都由人再看一遍,确认了才发给你。An automated pass goes over the whole file first: whether the dates line up, whether the coursework supports the research being claimed, whether each experience has anything to back it, whether the awards appear on public lists, whether the files themselves look odd, and whether any Common App limit is exceeded. Then a person reads every single finding before the report goes out.

13
个自动检测器automated detectors
33
份竞赛与夏校官方核验档案official verification profiles
100%
发现都经人工看过of findings human-reviewed