nsfc-research-content-writer

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

NSFC(二)研究内容编排写作器

NSFC (II) Research Content Organization & Writing Tool

目标输出(契约)

Target Output (Contract)

  • 写入落点(3 个文件)
    • extraTex/2.1.研究内容.tex
    • extraTex/2.2.特色与创新.tex
    • extraTex/2.3.年度研究计划.tex
  • 禁止改动
    main.tex
    extraTex/@config.tex
    、任何
    .cls/.sty
  • 编排原则:先把
    2.1
    写成“可验证闭环”,再从
    2.1
    抽取创新点生成
    2.2
    ,最后把
    2.1
    的任务拆分成三年里程碑生成
    2.3
  • Write Destinations (3 Files):
    • extraTex/2.1.Research Content.tex
    • extraTex/2.2.Characteristics and Innovations.tex
    • extraTex/2.3.Annual Research Plan.tex
  • Prohibited Modifications:
    main.tex
    ,
    extraTex/@config.tex
    , any
    .cls/.sty
    files
  • Organization Principle: First write Section 2.1 as a "verifiable closed loop", then extract innovation points from Section 2.1 to generate Section 2.2, and finally split the tasks in Section 2.1 into three-year milestones to generate Section 2.3.

参数与输出模式(建议显式提供)

Parameters and Output Modes (Recommended to Provide Explicitly)

  • project_root
    :标书项目根目录(如
    projects/NSFC_Young
  • output_mode
    (默认
    apply
    ):
    • preview
      :不直接写入文件;输出三段可复制粘贴的 LaTeX 正文草稿,并标注应写入的目标文件路径
    • apply
      :仅写入三份目标文件(见“目标输出”),不触碰其他文件
  • project_root
    : Root directory of the grant proposal project (e.g.,
    projects/NSFC_Young
    )
  • output_mode
    (default
    apply
    ):
    • preview
      : Do not write directly to files; output three draft LaTeX main text sections that can be copied and pasted, and mark the target file paths they should be written to
    • apply
      : Only write to the three target files (see "Target Output"), do not touch other files

必需输入(最小信息表)

Required Inputs (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

写入安全约束(必须遵守)

Writing Security Constraints (Must Comply)

  1. 仅编辑三份
    extraTex/2.*.tex
    文件;不得修改
    main.tex
    extraTex/@config.tex
    、任何
    .cls/.sty
  2. 目标文件若已包含标题命令(如
    \\subsection{...}
    /
    \\subsubsection{...}
    ),只替换正文内容,不改标题与结构层级
  3. 信息不全时先提问补齐,不要用“看起来像真的”的细节硬写
  1. Only edit the three
    extraTex/2.*.tex
    files; do not modify
    main.tex
    ,
    extraTex/@config.tex
    , or any
    .cls/.sty
    files
  2. If the target file already contains title commands (e.g.,
    \\subsection{...}
    /
    \\subsubsection{...}
    ), only replace the main text content, do not modify the title or structural hierarchy
  3. If information is incomplete, ask for supplements first; do not force-write details that "look real"

工作流(按顺序执行)

Workflow (Execute in Order)

  1. 定位项目与目标文件:确认
    project_root
    ,读取并仅编辑三份
    extraTex/2.*.tex
    文件;如目标文件不存在,提示用户先初始化/拷贝模板项目。
  2. 固定“子目标三件套”:把目标拆成 3–4 个子目标(建议编号
    S1–S4
    便于后续回溯),并对每个子目标强制写清:
    • 指标(可判定/可验收)
    • 对照/基线(与谁比、怎么比)
    • 数据来源/验证方案(样本/实验体系/评估方法)
  3. 生成
    2.1 研究内容
    (以“问题→目标→内容→路线→验证”为主线):
    • 研究问题与总体目标(不超过 2 段)
    • 子目标列表(3–4 个,逐条带三件套)
    • 研究内容与任务分解(对齐子目标,避免“写一堆方法但看不出解决什么”)
    • 技术路线与验证口径(对照/消融/外部验证/泄漏防控/统计方法)
  4. 2.1
    抽取
    2.2 特色与创新
    • 1–3 条即可;每条用“相对坐标系”表达:与主流路线 A/B 相比,本项目在 X 上不同,预计带来 Y,可用 Z 验证。
    • 避免绝对化措辞(如“首次”“领先”);如确需使用,必须给出可核验证据或改写为可审稿的相对表述。
  5. 2.1
    推导
    2.3 年度研究计划
    (三年不跨年):
    • 每年:年度目标 → 关键任务 → 里程碑(可验收)→ 可交付成果(论文/数据/原型/规范/软件等)
    • 里程碑必须与子目标挂钩(否则评审会认为“计划与研究内容脱节”)
  6. 一致性校验
    • 检查
      2.2
      创新点是否能回溯到
      2.1
      的具体任务与验证;
    • 检查
      2.3
      里程碑是否覆盖全部子目标,且每年都有可交付物。
    • 术语口径对齐:研究对象/缩写/指标命名尽量与
      (一)立项依据
      (三)研究基础
      保持一致(如项目中已存在)
  1. Locate Project and Target Files: Confirm
    project_root
    , read and only edit the three
    extraTex/2.*.tex
    files; if the target files do not exist, prompt the user to initialize/copy the template project first.
  2. Fix the "Sub-Goal Triplet": Split the target into 3–4 sub-goals (recommended to number them S1–S4 for subsequent traceability), and for each sub-goal, clearly write:
    • Metrics (verifiable/acceptance-able)
    • Comparison/Baseline (compare with whom, how to compare)
    • Data Source/Verification Scheme (samples/experimental system/evaluation method)
  3. Generate Section 2.1 Research Content (following the main line of Problem → Objective → Content → Route → Verification):
    • Research Problems and Overall Objectives (no more than 2 paragraphs)
    • List of Sub-Goals (3–4 items, each with the triplet)
    • Research Content and Task Decomposition (align with sub-goals, avoid "writing a bunch of methods without showing what problems they solve")
    • Technical Route and Verification Criteria (comparison/ablation/external verification/leakage prevention/statistical methods)
  4. Extract Section 2.2 Characteristics and Innovations from Section 2.1:
    • 1–3 items are sufficient; each item should be expressed using a "relative coordinate system": Compared with mainstream routes A/B, this project differs in X, is expected to bring Y, and can be verified with Z.
    • Avoid absolute wording (e.g., "first time", "leading"); if it is necessary to use such wording, provide verifiable evidence or rewrite it into a reviewable relative statement.
  5. Derive Section 2.3 Annual Research Plan from Section 2.1 (three years, no cross-year arrangements):
    • For each year: Annual Objective → Key Tasks → Milestones (verifiable) → Deliverables (papers/data/prototypes/specifications/software, etc.)
    • Milestones must be linked to sub-goals (otherwise reviewers will think "the plan is disconnected from the research content")
  6. Consistency Check:
    • Check whether the innovation points in Section 2.2 can be traced back to specific tasks and verification in Section 2.1;
    • Check whether the milestones in Section 2.3 cover all sub-goals, and there are deliverables every year.
    • Terminology Alignment: Try to keep the research objects/abbreviations/metric names consistent with Section (I) Project Justification and Section (III) Research Basis (if they already exist in the proposal)

验收标准(Definition of Done)

Acceptance Criteria (Definition of Done)

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

写作小抄(可选)

Writing Cheat Sheet (Optional)

  • 子目标“三件套”示例:references/subgoal_triplet_examples.md
  • 创新点“相对坐标系”示例:references/relative_coordinate_examples.md
  • 年度计划模板(确保里程碑可验收):references/yearly_plan_template.md
  • 三个输出文件的最小结构骨架(可复制粘贴):references/output_skeletons.md
  • 常见写作反模式与改写:references/anti_patterns.md
  • 验证口径菜单(对照/消融/外部验证/统计/泄漏防控):references/validation_menu.md
  • 术语口径对齐表(跨章节一致):references/terminology_sheet.md
  • Examples of Sub-Goal Triplet: references/subgoal_triplet_examples.md
  • Examples of Innovation Point "Relative Coordinate System": references/relative_coordinate_examples.md
  • Annual Plan Template (Ensure Milestones are Verifiable): references/yearly_plan_template.md
  • Minimum Structure Skeleton for Three Output Files (Copy-Pasteable): references/output_skeletons.md
  • Common Writing Anti-Patterns and Rewrites: references/anti_patterns.md
  • Verification Criteria Menu (Comparison/Ablation/External Verification/Statistics/Leakage Prevention): references/validation_menu.md
  • Terminology Alignment Table (Consistent Across Chapters): references/terminology_sheet.md

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
  • This tool does not maintain change history in this document; all changes are uniformly recorded in the root-level
    CHANGELOG.md
    .