# Prompt: Staff Certification Renewal Tracker > A copy-paste prompt that turns a pasted staff certification list into a renewal-tracking table with urgency tiers and verification gaps. ## Install Copy the content below into your project: # Prompt: Staff Certification Renewal Tracker A copy-paste prompt that turns a pasted staff certification list into a renewal-tracking table with urgency tiers and verification gaps. ## Start here Use this prompt when you already have a rough list of staff certifications and expiry dates (in an email, notes file or spreadsheet column) and need it turned into an organized renewal-tracking table. **What to paste:** the whole prompt below, then your own certification list as a separate message. If you have one, include the reference date you want the urgency tiers measured from (for example, today's date). If you don't add one, the prompt deliberately marks the table as blocked on the reference date instead of guessing. **Where to paste it:** any ordinary AI chat that accepts text. No terminal, account setup or API key is needed. **How to check the output:** the reply should have four parts — A: the table, B: verification gaps, C: counts, D: a manual spreadsheet procedure. Before you trust it, confirm that every entry you pasted appears exactly once, that no urgency tier is attached to a date you never supplied, and that Part C's counts match what you see in the table. ## What this prompt does It organizes, it does not complete. It reproduces entries faithfully, normalizes dates only when the original is unambiguous, assigns one of six fixed urgency tiers (Expired, Tier 1 critical 0-30 days, Tier 2 soon 31-60, Tier 3 plan 61-90, Tier 4 later, Unknown), and separates anything uncertain into a verification-gaps list. It never invents a completion date, renewal status, manager, cost or reminder. Ambiguous dates such as 03/04/2026 are kept as written and flagged, not interpreted. ## Inputs and limits Each entry should carry a name, a certification, and ideally an expiry date, plus any reference, role or note you have. Dates may be missing, partial or unevenly formatted; duplicates are expected and get reported rather than merged. Limits worth knowing before you paste anything: - Empty or unrecognizable input should make the assistant stop, not produce a sample table. - The prompt does not send reminders, update systems or schedule anything. It only prepares a draft for human review. - Anything you paste goes to whichever AI service you use, so treat the list as you would any confidential file and remove details you are not comfortable sharing. - Regulatory or legal wording in your list is only echoed back, never interpreted. The prompt contains a fictional worked example (Dana Okafor, Priya Raman, Sam Lee). Those people and dates are illustrative only — not real data and not test results. ## FAQ **Why doesn't it just convert every date?** Because some formats are genuinely two-way ambiguous, and a confident-looking wrong date is worse than a flagged one. The ambiguity stays visible in Part B so a person can resolve it. **Can I use it without a reference date?** Yes, but the whole table is then marked blocked on the reference date. Add one when you want urgency tiers. ## Verification note Source reviewed; runtime not tested. The prompt text and its intent were checked against the supplied material; no run of this prompt against a real certification list was performed here. ## Source and thanks Original TokRepo prompt (CC BY 4.0). Reference context: [ChatGPT release notes](), reviewed 2026-10-04. The complete prompt, its Chinese translation and the license are appended separately. ## Complete reusable prompt You are helping me build a renewal-tracking table for staff certifications. I will paste a raw list of people and their certifications. Your job is to organize it, not to complete it, and not to assume anything that is not in my input. INPUT I WILL PROVIDE A pasted block containing, per line or per short entry: a person's name, a certification or license name, and (if known) an expiry date, plus optionally a reference or ID, a required-by role, or a note. Dates may be missing, partial, or written in different formats. There may be duplicates or the same person listed twice. WHAT TO DO 1. Reproduce every entry faithfully in a table. Do not drop rows. Do not merge two different certifications for the same person unless they are truly the same certification, in which case note the duplication. 2. Normalize dates only when the original is unambiguous. If a date could be read two ways (for example 03/04/2026), keep the original string and flag it as ambiguous rather than choosing an interpretation. 3. For each row, assign an urgency tier based on the expiry date relative to a reference date I provide. If I do not provide a reference date, use no tier and mark the whole table as blocked on the reference date instead of guessing. 4. Compute days remaining only for rows with a complete, unambiguous expiry date. Otherwise leave the days-remaining cell as 'needs date'. 5. Produce a separate verification-gaps list: every row with a missing, partial, or ambiguous date; every duplicate; every certification name that appears in more than one spelling; every entry where the person is unclear. 6. Never invent a completion date, a renewal status, a manager, a cost, or a reminder. The table tracks what exists; it does not record what was done. URGENCY TIERS (use these labels, do not add others) - Expired: expiry date is before the reference date. - Tier 1 (critical): 0-30 days remaining. - Tier 2 (soon): 31-60 days remaining. - Tier 3 (plan): 61-90 days remaining. - Tier 4 (later): more than 90 days remaining. - Unknown: cannot be tiered because the date is missing or ambiguous. OUTPUT FORMAT Part A - Renewal-tracking table with these columns, in this order: Person | Certification | Raw expiry (as given) | Normalized date | Days remaining | Urgency tier | Source row note Part B - Verification gaps, grouped as: - Missing or partial dates - Ambiguous dates (show the original string and why it is ambiguous) - Duplicate or near-duplicate entries - Name or certification spelling inconsistencies - Rows I should confirm before anyone is contacted Part C - Counts: total rows, rows with complete dates, rows tiered, rows unknown, duplicates detected. Counts must match the table; if they do not, say so. Part D - Manual alternative: a short numbered procedure a person can follow in a spreadsheet or on paper to reproduce Parts A and B without any AI tool, including how to sort by days remaining once dates are complete. BOUNDARIES - You are preparing a tracking draft for human review. You are not sending reminders, not updating any system, and not scheduling anything. - If my input is empty or contains no recognizable certification lines, say so and stop rather than producing an example table. - If an entry mentions a regulatory or legal requirement, note only that the entry says so; do not interpret what the law requires. CHECK BEFORE YOU FINISH - Confirm every input entry appears exactly once in Part A or is listed as a duplicate in Part B. - Confirm no tiered row contains a date you supplied yourself. - Confirm each ambiguity flag quotes the original string from my input. - Confirm Parts C counts equal the actual table and gap-list contents. - If any check fails, state which check failed and fix the output before presenting it. FICTIONAL EXAMPLE INPUT Reference date: 2026-11-01 Dana Okafor - Food Handler Card - 12/2026 Dana Okafor - Forklift Operator - 2026-11-20 Priya Raman - CPR - 03/04/2026 Sam Lee - First Aid - (no date) Sam Lee - First Aid EXAMPLE OUTPUT SHAPE Part A: Dana Okafor | Food Handler Card | 12/2026 | (partial) | needs date | Unknown | month only; day unknown. Dana Okafor | Forklift Operator | 2026-11-20 | 2026-11-20 | 19 | Tier 1 (critical) | none. Priya Raman | CPR | 03/04/2026 | (kept as given) | needs date | Unknown | day/month order ambiguous. Sam Lee | First Aid | (none) | needs date | needs date | Unknown | date absent; duplicate entry. Part B: Missing dates - Sam Lee First Aid. Ambiguous - 03/04/2026 for Priya Raman. Duplicates - Sam Lee First Aid appears twice. Part C: total rows 5; complete dates 1; tiered 1; unknown 4; duplicates 1. Part D: numbered spreadsheet procedure. This worked example is illustrative only and represents no real person. Now wait for my input and apply the same rules to it. ## References and reuse - [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes) · Reviewed 2026-10-04 Original TokRepo prompt · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Reference documents retain their own rights. --- # 提示词：员工资质续期跟踪表 一段可直接粘贴的提示词，把员工资质清单整理成带紧急度分级和待核实清单的续期跟踪表。 ## 开始使用 当你手上已有一份粗略的员工资质与到期日期清单（在邮件、备忘文件或表格某一列里），想把它整理成一张续期跟踪表时，就用这段提示词。 **粘贴什么：** 先粘贴下面的完整提示词，然后另发一条消息粘贴你自己的资质清单。如果你有指定的参考日期（例如今天），一并写上，分级就以它为准。如果你没有提供，提示词会明确把整张表标记为“缺少参考日期”，而不是自行猜测。 **粘贴到哪里：** 任何接受文本的普通 AI 对话窗口。不需要终端、账号配置或 API 密钥。 **如何检查输出：** 回复应包含四个部分——A：表格，B：待核实清单，C：计数，D：手工表格操作步骤。在采信之前，请确认你粘贴的每一条都恰好出现一次，没有任何一行的紧急度分级基于你未提供的日期，且 C 部分的计数与表格实际内容一致。 ## 这段提示词做什么 它只做整理，不做补全。它忠实还原每一条记录，只在日期本身无歧义时才做规范化，给出六个固定分级之一（已过期、一级紧急 0-30 天、二级临近 31-60 天、三级计划 61-90 天、四级较晚、未知），并把所有不确定项单列成待核实清单。它绝不编造完成日期、续期状态、主管、费用或提醒。像 03/04/2026 这种有歧义的日期会原样保留并标记，不会被擅自解读。 ## 输入与限制 每条记录最好包含姓名、资质名称、到期日期，以及你手头有的编号、所属岗位或备注。日期可能缺失、不完整或格式不统一；重复条目是正常情况，会被报告而不会被合并。 粘贴前需要知道的限制： - 输入为空或无法识别时，助手应直接说明并停止，而不是给出示例表格。 - 提示词不会发送提醒、不会更新任何系统、也不会安排任何日程，只生成供人工复核的草稿。 - 你粘贴的内容会发送到你所用的 AI 服务，请按处理机密文件的方式对待这份清单，删除不便分享的细节。 - 清单中涉及的法规或法律表述只会被原样引用，不会被解读。 提示词里包含一个虚构示例（Dana Okafor、Priya Raman、Sam Lee）。这些人和日期仅作说明，不是真实数据，也不是测试结果。 ## 常见问题 **为什么不把所有日期都直接转换？** 因为有些格式确实存在两种合理读法，一个看起来确信但实际错误的日期比一个被标记出来的日期更糟。歧义会保留在 B 部分，交由人来确认。 **没有参考日期能用吗？** 可以，但整张表会被标记为缺少参考日期。想要紧急度分级时就补上它。 ## 核实说明 已审阅来源，未做运行时测试。提示词文本及其意图已对照所给材料检查；此处未用真实资质清单实际运行该提示词。 ## 来源与致谢 TokRepo 原创提示词（CC BY 4.0）。参考背景：[ChatGPT release notes]()，审阅于 2026-10-04。完整提示词、中文译文与许可协议另行附上。 ## 完整可复制提示词 你正在帮助我构建一张员工资质续期跟踪表。我会粘贴一份包含人员和其资质的原始清单。你的任务是整理它，而不是补全它，并且不要假设任何我的输入中没有的信息。 我将提供的输入 一个粘贴的文本块，按行或按短条目包含：人员姓名、资质或许可证名称，以及（如果已知）到期日期，另外可选地包含编号或 ID、所需岗位或备注。日期可能缺失、不完整，或以不同格式书写。可能存在重复项，或同一个人被列出两次。 要做什么 1. 在表格中忠实还原每一条记录。不要丢弃行。不要把同一个人的两项不同资质合并，除非它们确实是同一项资质，此时应注明该重复。 2. 只在原始日期无歧义时才做规范化。如果日期有两种读法（例如 03/04/2026），保留原始字符串并将其标记为有歧义，而不是选择一种解读。 3. 对于每一行，根据到期日期相对于我提供的参考日期，分配一个紧急度分级。如果我没有提供参考日期，则不要分配分级，而是将整张表标记为“缺少参考日期”，而不是自行猜测。 4. 仅对具有完整、无歧义到期日期的行计算剩余天数。否则将剩余天数列留为“需要日期”。 5. 生成单独的待核实清单：每一行日期缺失、不完整或有歧义的记录；每一个重复项；每一个出现多种拼写的资质名称；每一条人员不明确的记录。 6. 绝不编造完成日期、续期状态、主管、费用或提醒。表格跟踪的是现有信息；它不记录已完成的事项。 紧急度分级（使用这些标签，不要添加其他） - 已过期：到期日期早于参考日期。 - 一级（紧急）：剩余 0-30 天。 - 二级（临近）：剩余 31-60 天。 - 三级（计划）：剩余 61-90 天。 - 四级（较晚）：剩余超过 90 天。 - 未知：因日期缺失或有歧义而无法分级。 输出格式 A 部分 - 续期跟踪表，按此顺序包含以下列： 人员 | 资质 | 原始到期日期（按给定） | 规范化日期 | 剩余天数 | 紧急度分级 | 来源行备注 B 部分 - 待核实清单，分组如下： - 缺失或不完整的日期 - 有歧义的日期（展示原始字符串及为何有歧义） - 重复或近似重复的条目 - 姓名或资质拼写不一致 - 在联系任何人之前需要我确认的行 C 部分 - 计数：总行数、具有完整日期的行数、已分级的行数、未知的行数、检测到的重复项数。计数必须与表格一致；如果不一致，请说明。 D 部分 - 手工替代方案：一段简短的分步流程，供人在不使用任何 AI 工具的情况下，在电子表格或纸上复现 A 部分和 B 部分，包括一旦日期完整后如何按剩余天数排序。 边界 - 你正在为人工审核准备一份跟踪草稿。你不是在发送提醒、不是在更新任何系统，也不是在安排任何日程。 - 如果我的输入为空或不包含可识别的资质行，请直接说明并停止，而不是给出示例表格。 - 如果某条记录提到法规或法律要求，只注明该记录如此陈述；不要解读法律要求什么。 完成前检查 - 确认每一条输入记录在 A 部分恰好出现一次，或在 B 部分被列为重复项。 - 确认没有任何已分级的行包含你自己提供的日期。 - 确认每个歧义标记都引用了我输入中的原始字符串。 - 确认 C 部分的计数等于表格和清单的实际内容。 - 如果任何检查未通过，说明是哪项检查未通过，并在呈现输出前修正。 虚构示例输入 参考日期：2026-11-01 Dana Okafor - 食品处理人员证 - 12/2026 Dana Okafor - 叉车操作员 - 2026-11-20 Priya Raman - 心肺复苏 - 03/04/2026 Sam Lee - 急救 - （无日期） Sam Lee - 急救 示例输出结构 A 部分：Dana Okafor | 食品处理人员证 | 12/2026 | （部分）| 需要日期 | 未知 | 仅月份；日未知。Dana Okafor | 叉车操作员 | 2026-11-20 | 2026-11-20 | 19 | 一级（紧急）| 无。Priya Raman | 心肺复苏 | 03/04/2026 | （原样保留）| 需要日期 | 未知 | 日/月顺序有歧义。Sam Lee | 急救 | （无）| 需要日期 | 需要日期 | 未知 | 日期缺失；重复条目。 B 部分：缺失日期——Sam Lee 急救。歧义——Priya Raman 的 03/04/2026。重复——Sam Lee 急救出现两次。 C 部分：总行数 5；完整日期 1；已分级的 1；未知的 4；重复 1。 D 部分：编号的电子表格操作步骤。 此示例仅作说明之用，不代表任何真实人物。现在请等待我的输入，并对它应用同样的规则。 ## 参考资料与复用 - [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes) · Reviewed 2026-10-04 TokRepo 原创提示词 · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)。参考资料保留各自原有权利。 --- Source: https://tokrepo.com/en/workflows/prompt-staff-certification-renewal-tracker-2373b945 Author: Prompt Lab