Back to Details

recommendation-letter-writing

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Recommendation Letter Writing

推荐信撰写

Use When

适用场景

The user wants to draft, rewrite, or refine a graduate admissions recommendation letter.
用户需要起草、重写或优化研究生入学申请推荐信时使用。

Workflow

工作流程

Use the user's language for all conversation, questions, explanations, and review notes, but keep the actual recommendation letter output in English unless the user explicitly asks otherwise.
所有对话、提问、解释和审核说明使用用户的语言，但实际输出的推荐信内容默认保留英文，除非用户明确要求其他语言。

Phase 1: Collect Information

第一阶段：信息收集

Load ONLY references/info-requirements.md, then guide the user with that checklist and its material-mining strategy.
  • Show the checklist so the user can provide everything in one pass if they want.
  • After receiving input, identify what is already covered and ask follow-up questions only for missing required categories.
  • If story material is thin, ask targeted follow-up questions using the mining strategy.
  • Use careful inference only as a questioning aid. Suggest likely scenes in question form, keep only what the user confirms, and discard what they reject or skip.
  • Optional categories should be asked only when useful.
⛔ Do not proceed to Phase 2 until every required category in references/info-requirements.md has been addressed. A category is addressed only when the user has either provided all required sub-fields or explicitly confirmed they have nothing more for that category. See the Validation Rules section in that reference.
仅加载 references/info-requirements.md，然后按照该文件中的 Checklist 和素材挖掘策略引导用户提供信息。
  • 向用户展示 Checklist，方便用户可以一次性提供所有所需材料。
  • 收到用户输入后，确认已收集到的内容，仅针对缺失的必填类别追加提问。
  • 如果故事素材不足，按照挖掘策略提出有针对性的跟进问题。
  • 仅将合理推断作为提问辅助，以问题形式给出可能的场景，仅保留用户确认的内容，舍弃用户拒绝或跳过的内容。
  • 仅在有必要时询问可选类别的信息。
⛔ 在 references/info-requirements.md 中所有必填类别都处理完成前，不得进入第二阶段。仅当用户已提供某类别下所有必填子字段，或明确确认该类别无更多信息可提供时，才算该类别处理完成。详见该参考文件中的验证规则部分。

Phase 2: Confirm Information and Judge Sufficiency

第二阶段：信息确认与充分性判定

  1. Summarize the collected information and ask the user to confirm accuracy and missing items.
  2. Evaluate whether the material is sufficient using references/info-requirements.md.
  3. Tell the user clearly whether the material is sufficient or thin, and explain why.
⛔ Do not proceed to Phase 3 until the user explicitly confirms the summary is accurate.
  1. 汇总已收集的信息，请用户确认信息准确性以及是否有遗漏。
  2. 参考 references/info-requirements.md 评估素材是否足够。
  3. 明确告知用户素材是充分还是不足，并解释原因。
⛔ 在用户明确确认汇总信息准确之前，不得进入第三阶段。

Phase 3: Produce the Letter

第三阶段：生成推荐信

Branch A: Sufficient Material -> Final Version

分支A：素材充分 -> 终版文件

Follow this pipeline in order:
text
Step 1: Distill source material
    ↓
Step 2: Build a structured outline
    ↓
Step 3: Write the recommendation letter
    ↓
Step 4: Run an independent quality check
按照以下流程顺序执行：
text
Step 1: 提炼源素材
    ↓
Step 2: 搭建结构化大纲
    ↓
Step 3: 撰写推荐信
    ↓
Step 4: 执行独立质量检查
Step 1: Distill source material
步骤1：提炼源素材
Load references/writing-instructions.md and follow its workflow for distilling the strongest observable stories.
加载 references/writing-instructions.md，按照其工作流提炼最有说服力的真实案例。
Step 2: Build the outline
步骤2：搭建大纲
Use the outline structure defined in references/writing-instructions.md.
使用 references/writing-instructions.md 中定义的大纲结构。
Step 3: Write the recommendation letter
步骤3：撰写推荐信
Load references/output-format.md and follow its structured Markdown sections.
加载 references/output-format.md，遵循其规定的Markdown结构化章节撰写。
Step 4: Quality check
步骤4：质量检查
Load references/quality-checklist.md and run the checklist item by item.
  • Group results by severity: Error / Warning / Pass.
  • If there are Error-level issues, fix them before returning the revised version.
加载 references/quality-checklist.md，逐条对照检查。
  • 按严重程度分组展示结果：错误 / 警告 / 通过。
  • 如果存在错误级问题，先修复问题再返回修订版本。

Branch B: Thin Material -> Reference Draft + Follow-up Checklist

分支B：素材不足 -> 参考草稿 + 跟进清单

  • Use the workflow in references/writing-instructions.md to write a complete recommendation letter from the available facts plus careful inference.
  • Do not label inferred details inline in the letter body.
  • After the letter, append an 
    AI Notes
    section listing what was inferred and what the user should verify or replace.
  • Output a follow-up checklist of the specific details still worth collecting.
  • On the next conversation, the user can bring back the reference draft plus new material.
  • 参考 references/writing-instructions.md 中的工作流，基于现有事实加上合理推断撰写完整的推荐信。
  • 不要在推荐信正文中标记推断出的细节。
  • 在推荐信末尾添加 
    AI Notes
    板块，列出哪些内容是推断得出的，以及用户需要核实或替换的内容。
  • 输出后续仍值得收集的具体细节的跟进清单。
  • 下次对话时，用户可以提供参考草稿+新素材继续优化。

Output Format

输出格式

  • Main body: Markdown.
  • Optional translation: if the user's working language is Chinese, or they explicitly ask for it, append a Chinese translation.
  • Reference mode only: append an 
    AI Notes
    section after the letter.
  • Reference mode only: append the follow-up checklist of missing details.
  • Quality check: group results as Error / Warning / Pass.
  • Output ordering: follow references/output-format.md.
  • 正文：Markdown格式。
  • 可选翻译：如果用户的工作语言是中文，或用户明确要求翻译，可在末尾附加中文翻译。
  • 仅参考模式：在推荐信末尾附加 
    AI Notes
    板块。
  • 仅参考模式：附加缺失细节的跟进清单。
  • 质量检查：按错误 / 警告 / 通过分组展示结果。
  • 输出顺序：遵循 references/output-format.md 的规定。