nsfc-research-foundation-writer

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

NSFC(三)研究基础编排写作器

NSFC (3) Research Basis Organization Writer

目标输出(契约)

Target Output (Contract)

  • 写入落点(2 个文件)
    • extraTex/3.1.研究基础.tex
      (包含“研究风险的应对措施”)
    • extraTex/3.2.工作条件.tex
  • 禁止改动
    main.tex
    extraTex/@config.tex
    、任何
    .cls/.sty
  • 核心目标:用“证据链 + 条件对位 + 风险预案”回答评审的三个问题:你做过吗?你做得成吗?出问题你怎么兜底?
  • Output Destinations (2 files)
    • extraTex/3.1.Research Basis.tex
      (includes "Response Measures for Research Risks")
    • extraTex/3.2.Working Conditions.tex
  • Prohibited Modifications
    main.tex
    ,
    extraTex/@config.tex
    , any
    .cls/.sty
    files
  • Core Objective:Answer three reviewer questions with "evidence chain + condition alignment + risk contingency plan": Have you done similar work? Can you complete this project? How will you handle issues if they arise?

必需输入(最小信息表)

Required Input (Minimum Information Form)

  • 若用户未提供,请先收集/补全:references/info_form.md
  • If not provided by the user, collect/supplement first: references/info_form.md

工作流(按顺序执行)

Workflow (Execute in Order)

  1. 定位项目与目标文件
    • 验证
      project_root
      是否存在,不存在时报错并提示用户指定正确路径
    • 检查
      extraTex/
      目录是否存在,不存在时尝试创建或报错提示
    • 确认
      project_root
      output_mode
      (默认为
      apply
      • preview
        模式:只生成内容预览,不写入文件(适合调试)
      • apply
        模式:将生成的内容写入
        extraTex/3.1.研究基础.tex
        extraTex/3.2.工作条件.tex
    • 仅编辑这两个文件,禁止修改
      main.tex
      extraTex/@config.tex
      、任何
      .cls/.sty
  2. 生成
    3.1 研究基础
    (证据链优先)
    • 研究积累:围绕
      2.1
      的关键任务,列出“做过什么/掌握什么/已有平台什么”。
    • 阶段性成果:只写可核验内容(论文/专利/数据/原型/预实验现象);不确定的细节用占位符要求用户补齐。
    • 可行性四维:理论/技术/团队/条件各给 1–3 个支撑点,并与研究内容逐条对齐。
  3. 3.1
    中显式写“研究风险的应对措施”
    • 至少 3 条风险(技术/进度/资源各至少 1 条)
    • 每条:风险描述 → 早期信号(触发阈值/现象)→ 预案/替代路线(含降级目标与可交付)
  4. 生成
    3.2 工作条件
    (条件对位研究内容)
    • 已具备条件:逐条列出,格式建议:
      • 平台:XXX 平台(已具备 / 可访问)
      • 数据:XXX 数据集(已获取 / 可公开获取)
      • 样本:XXX 医院/机构(已签署合作协议 / 伦理审批中)
      • 算力:XXX 服务器 / GPU(已配置 / 共享使用)
      • 团队分工:成员 A 负责 XX,成员 B 负责 YY
      • 合规路径:伦理审批(XXX 委员会,周期 X 个月)
    • 尚缺条件与解决途径:逐条列出,格式建议:
      • 缺少条件:XXX(影响:YYY)
      • 解决途径:采购 / 合作 / 替代数据源 / 实验降级方案
      • 时间表与责任人:如用户提供,应写入;如未提供,用占位符标记
  5. 一致性校验
    • 校验 1:检查
      3.2 工作条件
      是否能逐条支撑
      2.1
      的关键任务
      • 方法:列出
        2.1
        的每个关键任务,确认
        3.2
        中有对应的条件支撑
      • 示例:
        • 任务 1:XXX 实验 → 条件:XXX 平台、XXX 样本
        • 任务 2:YYY 算法 → 条件:YYY 算力、ZZZ 数据
    • 校验 2:检查风险预案是否与年度计划可兼容
      • 方法:列出
        2.3
        的每年里程碑,确认风险预案有对应的降级方案
      • 示例:
        • 第一年:样本获取 → 风险:样本入组慢 → 预案:有替代数据源与降级验证方案
    • 校验失败时的处理
      • 如果发现不一致,标记为
        [需补充:XXX]
        并提示用户
      • 如果用户未提供
        2.1
        2.3
        ,跳过对应校验并提示用户
  1. Locate Project and Target Files
    • Verify if
      project_root
      exists; if not, report an error and prompt the user to specify the correct path
    • Check if the
      extraTex/
      directory exists; if not, attempt to create it or report an error
    • Confirm
      project_root
      and
      output_mode
      (default is
      apply
      )
      • preview
        mode: Generate only content preview without writing to files (suitable for debugging)
      • apply
        mode: Write generated content to
        extraTex/3.1.Research Basis.tex
        and
        extraTex/3.2.Working Conditions.tex
    • Only edit these two files; prohibited from modifying
      main.tex
      ,
      extraTex/@config.tex
      , any
      .cls/.sty
      files
  2. Generate
    3.1 Research Basis
    (Evidence Chain Priority)
    • Research Accumulation: Focus on the key tasks in
      2.1
      , list "what you have done/what you have mastered/what platforms you already have".
    • Phased Achievements: Only write verifiable content (papers/patents/data/prototypes/preliminary experiment phenomena); use placeholders to request the user to supplement uncertain details.
    • Four Dimensions of Feasibility: Provide 1–3 supporting points for each of theory/technology/team/conditions, and align them one by one with the research content.
  3. Explicitly write "Response Measures for Research Risks" in
    3.1
    • At least 3 risks (at least 1 each for technology/schedule/resources)
    • For each risk: Risk Description → Early Warning Signs (trigger threshold/phenomenon) → Contingency Plan/Alternative Route (including downgraded objectives and deliverables)
  4. Generate
    3.2 Working Conditions
    (Conditions Align with Research Content)
    • Existing Conditions: List item by item, recommended format:
      • Platform: XXX Platform (Available / Accessible)
      • Data: XXX Dataset (Obtained / Publicly Accessible)
      • Samples: XXX Hospital/Institution (Cooperation Agreement Signed / Ethical Review in Progress)
      • Computing Power: XXX Server / GPU (Configured / Shared Use)
      • Team Division of Labor: Member A is responsible for XX, Member B is responsible for YY
      • Compliance Path: Ethical Review (XXX Committee, Cycle X Months)
    • Missing Conditions and Solutions: List item by item, recommended format:
      • Missing Condition: XXX (Impact: YYY)
      • Solution: Procurement / Cooperation / Alternative Data Source / Experiment Downgrade Plan
      • Timeline and Responsible Person: Write it if provided by the user; use a placeholder if not provided
  5. Consistency Check
    • Check 1: Verify if
      3.2 Working Conditions
      can support each key task in
      2.1
      one by one
      • Method: List each key task in
        2.1
        , confirm that corresponding supporting conditions exist in
        3.2
      • Example:
        • Task 1: XXX Experiment → Conditions: XXX Platform, XXX Samples
        • Task 2: YYY Algorithm → Conditions: YYY Computing Power, ZZZ Data
    • Check 2: Verify if risk contingency plans are compatible with the annual plan
      • Method: List each annual milestone in
        2.3
        , confirm that the risk contingency plans have corresponding downgrade solutions
      • Example:
        • Year 1: Sample Acquisition → Risk: Slow Sample Recruitment → Contingency Plan: Alternative Data Source and Downgraded Verification Plan
    • Handling Check Failures:
      • If inconsistencies are found, mark them as
        [Need to Supplement: XXX]
        and prompt the user
      • If the user does not provide
        2.1
        or
        2.3
        , skip the corresponding check and prompt the user

验收标准(Definition of Done)

Acceptance Criteria (Definition of Done)

质量契约验证(来自 config.yaml)

Quality Contract Verification (from config.yaml)

  • 研究基础必须覆盖:
    • 与本项目相关的研究积累(证据链)
    • 已取得的阶段性成果(可核验)
    • 可行性要点(理论/技术/团队/条件)
    • 研究风险与应对措施(显式写出)
  • 工作条件必须覆盖:
    • 已具备条件(平台/数据/样本/算力/团队/合规)
    • 尚缺条件与解决途径(时间/预算/合作/采购/替代方案)
  • Research Basis must cover:
    • Research accumulation related to this project (evidence chain)
    • Achieved phased results (verifiable)
    • Feasibility key points (theory/technology/team/conditions)
    • Research risks and response measures (explicitly written)
  • Working Conditions must cover:
    • Existing conditions (platform/data/samples/computing power/team/compliance)
    • Missing conditions and solutions (time/budget/cooperation/procurement/alternative plans)

详细检查清单

Detailed Checklist

见:references/dod_checklist.md
See: references/dod_checklist.md

常见问题与边缘情况

Common Questions and Edge Cases

Q1: 项目目录中没有
extraTex/
目录怎么办?

Q1: What if there is no
extraTex/
directory in the project directory?

A: 本技能会自动尝试创建
extraTex/
目录。如果失败,请手动创建:
bash
mkdir -p projects/NSFC_Young/extraTex
A: This skill will automatically attempt to create the
extraTex/
directory. If it fails, please create it manually:
bash
mkdir -p projects/NSFC_Young/extraTex

Q2: 我还没有写
2.1 研究内容
,可以直接写
3.1 研究基础
吗?

Q2: Can I directly write
3.1 Research Basis
without writing
2.1 Research Content
?

A: 不建议。
3.2 工作条件
需要与
2.1
的关键任务对齐。建议先使用
nsfc-research-content-writer
完成
2.1
A: Not recommended.
3.2 Working Conditions
needs to align with the key tasks in
2.1
. It is recommended to complete
2.1
using
nsfc-research-content-writer
first.

Q3: 信息表中的某些内容我不方便公开怎么办?

Q3: What if some content in the information form is inconvenient to disclose?

A: 可以提供"可核验线索"而非完整内容。例如:
  • ❌ "我们在 Nature 上发表了论文 XXX"
  • ✅ "我们在 Nature 上发表过相关论文(可提供审稿人推荐信)"
A: You can provide "verifiable clues" instead of complete content. For example:
  • ❌ "We published paper XXX in Nature"
  • ✅ "We have published relevant papers in Nature (can provide reviewer recommendation letters)"

Q4: 风险应对必须写 3 条吗?

Q4: Do I have to write 3 risk responses?

A: 是的,至少 3 条(技术/进度/资源各至少 1 条)。如果风险较少,可以写"风险较低,暂无其他明显风险"。
A: Yes, at least 3 (at least 1 each for technology/schedule/resources). If there are few risks, you can write "Risks are low, no other obvious risks for now".

Q5: 我的信息表内容很少,能生成完整的研究基础吗?

Q5: Can a complete research basis be generated if the information form has very little content?

A: 可以。AI 会根据你提供的信息生成内容,并在不确定的地方使用占位符(如
[请补充:XXX]
),请你后续补全。
A: Yes. AI will generate content based on the information you provide, and use placeholders (such as
[Please Supplement: XXX]
) in uncertain areas. Please supplement them later.

变更记录

Change Log

  • 本技能不在本文档内维护变更历史;统一记录在根级
    CHANGELOG.md
  • The change history of this skill is not maintained in this document; it is uniformly recorded in the root-level
    CHANGELOG.md
    .