一份合同覆盖学校名下多个 STS 申请：批量预检队列、逐申请审计、九关台账，加一份可直接转阅学校领导的协调人视图，全程只用申请人代号。每一份交付报告都含人工复核。One contract across the school's STS applications: a batch preflight queue, per-application audits, the nine-gate ledger, plus a coordinator view that can be forwarded to school leadership as-is — project codes only, throughout. Human review is included on every delivered report.
学校里同时有几个学生冲 STS 时，协调老师面对的不是一个申请的深度问题，而是横向问题：哪几个申请卡在同一个问题上、谁的受监管研究文档还没启动、谁的推荐轨道一条都没发出去、哪几项已经逾期。逐个打开报告翻一遍，既慢又容易漏。When several students at one school are applying to STS, the coordinator's problem is not depth on one application — it is breadth: which applications are stuck in the same hole, whose regulated-research documentation has not started, whose recommendation tracks have not been invited, and what is already overdue. Opening each report one by one is slow and easy to miss things in.
学校版在赛季版全部能力之上加了两件事：批量，按申请人名册逐个跑完整交付流；协调人视图，四节固定结构（横向风险表 / 逾期表 / 批量倒排 / 逐申请状态行），同一问题跨两个以上申请时自动合并成一行，让系统性问题一眼可见。The School tier adds two things on top of everything in the Season tier: batch — the full delivery flow run applicant by applicant down the roster; and the coordinator view — four fixed sections (cross-applicant risk table / overdue table / batch countdown / one status line per application), where an issue spanning two or more applications is merged into a single row so systemic problems are visible at a glance.
单个申请人自用请看学生赛季版；只想在提交前做一次外部把关请看提交预检版。For a single applicant, see the Student Season tier; for one external check before submission, see the Preflight Report.
学校同时支持几个申请人、老师需要一张能直接转阅校领导的状态表：选这一档。每个申请人都是完整三件套，批量不等于查得浅，每份报告都有人工复核过再交。Choose this tier when a school supports several applicants at once and the teacher needs a status view that can go straight to school leadership. Every applicant gets the full three-report set; batch doesn’t mean shallow, and every report is human-reviewed before delivery.
只有一个申请人就不用上学校版：提交前查一次选预检，整赛季跟进选赛季版。学校流程必须依赖托管模式的，当前版本满足不了：托管模式在隐私、保留与法务工作完成前不提供。One applicant? Don’t buy the school tier — a single pre-submission check is the Preflight, season-long support is the Student Season tier. And if your school’s process requires a hosted mode, this version can’t serve it: hosted mode is not offered until the privacy, retention and legal work is done.
名册上的每个申请都按完整交付流跑：M1/M2/M3 三件套缺一即未完成，三格式（HTML / PDF / DOCX）交付，且每一份都经人工复核后才交付。Every application on the roster runs the full delivery flow: the M1/M2/M3 set (missing any one means incomplete), delivered in three formats (HTML / PDF / DOCX), and every single one passes human review before delivery.
这份视图默认是给学校领导转阅的，因此它本来就不含任何身份信息；也不对申请人做"值得投入程度"排名，不输出任何入选预测。This view is designed to be forwarded to school leadership, which is exactly why it carries no identity information; it also never ranks applicants by how "worth investing in" they are and never outputs a selection prediction.
cohort-roster.schema.json 维护：仅代号与非敏感字段，不收姓名、出生日期、联系方式、学校地址The roster follows cohort-roster.schema.json: project codes and non-sensitive fields only — no names, birth dates, contact details, or school addresses这一档买的是按名册批量：名册上的申请人，一个个查过去，每人拿到的都是完整的三件套；名册、排队顺序和给老师的汇总表，由我们统一维护。批量不等于查得浅，只是多了一层排期和汇总。This tier runs as batch (multi-applicant): the full delivery flow is run application by application down the roster, with the roster, queue order, and coordinator view maintained by the cohort-management capability. Batch does not mean shallower review — each application still receives the complete core set; scheduling and the coordinator layer sit on top.
对照另外两档：预检是查一次，交付归档就封存；赛季版是改一版查一版，一个申请整赛季反复查。For contrast: once is the Preflight tier's semantics — one delivery, sealed on archiving; repeat is the Season tier's — one application audited repeatedly with M3 updated in place.
当前学校版按仅本地方式交付：名册与材料留在本机处理。托管式（hosted）学校模式的隐私、保留与法务工作尚未完成，未完成前不提供。这一条写在这里，是为了让贵校在评估阶段就知道边界在哪。The School tier is currently delivered local-only: the roster and materials stay on the local machine. Hosted school mode is not offered until its privacy, retention, and legal work is complete — stated here so your school knows the boundary while still evaluating.
本档人工复核是包含项，也是必经步骤：任何一份交付报告在人工填写复核块（复核人、复核日期、结论、修改闭环确认）之前都不算完成，也不得在协调人视图里标为已交付。复核块只由人工填写，系统绝不代填。这与预检、赛季版的"可选加购"是本质区别：学校版的报告会进入学校的档案与转阅链条，责任更重。At this tier human review is included and mandatory: no delivered report is complete until a person fills in the reviewer block (reviewer, review date, conclusion, change-closure confirmation), and it may not be marked delivered in the coordinator view before that. The block is filled by a person only — never auto-filled. This is the substantive difference from the optional add-on at Preflight and Season: School-tier reports enter a school's records and get forwarded, so the responsibility is heavier.
| 功能 / 服务Feature / service | 免费Free自己动手Do it yourself | 机构学校版School按名册批量Batch, all applicants |
|---|---|---|
| 规则讲解、材料模板、自查技能Rules, templates, self-check skills | ✓ | ✓ |
| 三份核心清单：内容问题、材料核对、整改时间表Core set: content audit, gap report, action plan | 自己跑Self-run | ✓ |
| 2027 规则逐条核对（注明出处，随官方更新）Rule-by-rule check, sourced and maintained | 参考文档Reference only | ✓ |
| 资格与受监管研究的分流核对Eligibility + regulated-research triage | 参考资料Reference only | ✓ |
| AI 使用与支持披露检查AI and support-disclosure review | 参考资料Reference only | ✓ |
| 推荐信、成绩单交没交（只记状态，不看正文）Recommendation / transcript status (status only) | 参考资料Reference only | ✓ |
| 申请进度台账（九个关口的证据记录）Nine-gate evidence ledger | — | ✓ |
| Finalist 面试练习（官方通知后）Finalist interview practice (after official notice) | 参考资料Reference only | ✓ |
| 七项按需支持（O1–O7）On-request menu (O1–O7) | 自己跑Self-run | ✓ |
| 多申请人名册、批量检查、老师汇总表Multi-applicant roster + coordinator view | — | ✓ |
| 人工复核（每份报告有人看过再交）Human review on every report | — | ✓ |
✓ 是我们来完成；「自己跑」是工具免费给你、自己动手；「加购」可选；— 不提供。四档的完整对照在总览页。✓ = we do it; “self-run” = free tools, your hands; add-on = optional paid; — = not offered. Full comparison on the overview page.
以及不论哪个档位、花多少钱都不提供的：代写 essay / 申请回答 / Research Report / 引用 / 推荐信；Scholar、Finalist 或获奖预测。这两条是红线，不是加钱能解锁的功能。And what no tier has at any price: ghostwritten essays / application answers / Research Report text / citations / recommendations, and Scholar, Finalist, or award prediction. These are red lines, not locked premium features.
不代写、不预测入选、推荐信只核状态、不替你确认事实、和主办方没有关系：五条规矩每一档都一样，全文在总览页。No ghostwriting, no selection prediction, recommendations tracked by status only, facts attested by the applicant, independent of the organizers — the same five rules in every tier; full text on the overview page.
在 ¥99,800–159,800 / 年之间，按申请人数议定，以双方确认的报价单为准。Agreed between ¥99,800 and ¥159,800 per year by roster size, fixed in the confirmed quote.
可以，它就是按这个用途设计的：全程只用申请人代号与角色名，不含任何身份信息，也不含推荐信或成绩单内容。Yes — it is designed for exactly that: project codes and role names only, with no identity information and no recommendation or transcript contents.
不会。当前学校版按仅本地方式交付，名册与材料留在本机处理，不用于训练模型。托管式学校模式在隐私、保留与法务工作完成前不提供。No. The School tier is delivered local-only: the roster and materials stay on the local machine and are never used to train models. Hosted school mode is not offered until its privacy, retention, and legal work is complete.
该申请会持续出现在逾期表里，负责角色与建议动作写清楚，交给协调人分派。我们不替学校做取舍，也不会因为材料不齐就给出任何关于该申请前景的判断。That application keeps appearing in the overdue table with its owner role and suggested action spelled out, for the coordinator to assign. We do not make the school's trade-offs for it, and incomplete materials never produce any judgment about that application's prospects.
不想写邮件的话，填几项基本信息就行，我们会主动联系你。No email needed — a few basics and we’ll reach out to you.