Agent skill

NSFC Grant Rationale Writer

by huangwb8 in huangwb8/ChineseResearchLaTeX

Writes, restructures, reviews and polishes the rationale section of NSFC research grant applications in LaTeX, with backups and a diff before every write.

MITAuto-check passedResearch & Science

SKILL.md written in Chinese; this summary is our English description.

Install NSFC Grant Rationale Writer

skills CLI
$ npx skills add huangwb8/ChineseResearchLaTeX --skill nsfc-justification-writer -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install huangwb8/ChineseResearchLaTeX nsfc-justification-writer --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/huangwb8/ChineseResearchLaTeX.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/nsfc-justification-writer .claude/skills/nsfc-justification-writer && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
nsfc-justification-writer
GitHub stars
2.9k
Token cost
~945 tokens
SKILL.md length
201 words
Files
79 (incl. scripts, references, assets)
Skills in repo
27
Repo updated
First seen
Licence
MIT

At a glance

Writes, restructures, reviews and polishes the rationale section of NSFC research grant applications in LaTeX, with backups and a diff before every write.

  • Works in 6 steps: 先写领域价值/必要性与已有证据,再将瓶颈转成待解释的关系、机制、性质或边界。 → 科学问题用疑问句追问认知缺口;假设用可证伪陈述句预测结果。研究目标、路线、验证终点… → 假设前以事实、文献和缺口为主;干预、比较、终点、路线放在假设后。外部文献的方法描述… → …
  • Drafting the research rationale section of an NSFC application in LaTeX
  • SKILL.md covers 流程 and 约束
  • Checking that a proposal's question, hypothesis and evidence line up

What it does

The target is the rationale section of a Chinese NSFC funding proposal: research significance, the state of the field, the scientific question and the hypothesis. The agent reads the LaTeX file (searching for `main.tex` when none is given), checks that the problem, hypothesis, evidence and research content hang together and that citation keys exist, then runs a semantic review that sorts findings into factual problems, logic problems and expression suggestions.

By default it works without stopping to ask you anything: it drafts the full text or a unified diff, confirms that structure commands and citations are unchanged, backs up the file, writes the change and shows the diff with a rollback path. `scripts/run.py` offers preview, refs, wordcount, diagnose, coach, review, diff and rollback commands, and `references/` holds guidelines on questions, hypotheses, readability and overclaiming. The skill text is in Chinese.

When your agent uses it

  • Drafting the research rationale section of an NSFC application in LaTeX
  • Checking that a proposal's question, hypothesis and evidence line up
  • Polishing dense sentences in a rationale section without changing facts or citations
  • Counting words or auditing citation keys in a proposal section

Example prompts

  • “Rewrite the rationale in ./proposal/main.tex so the scientific question and hypothesis are clearly separated.”
  • “Review the research significance section of my grant application and list factual, logic and wording problems.”
  • “Show me the diff you would apply to the rationale, then roll it back.”
  • “Check the word count and missing citation keys in the current rationale section.”

Requirements

  • Python, to run `scripts/run.py`
  • A LaTeX grant project with a main file such as `main.tex`

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

  1. 先写领域价值/必要性与已有证据,再将瓶颈转成待解释的关系、机制、性质或边界。
  2. 科学问题用疑问句追问认知缺口;假设用可证伪陈述句预测结果。研究目标、路线、验证终点不冒充问题/假设。
  3. 假设前以事实、文献和缺口为主;干预、比较、终点、路线放在假设后。外部文献的方法描述可作证据,不等于本项目方案前置。
  4. 切入点/贡献无需固定标题或段数;理论、混合、工程导向只改变措辞与证据重心,不是通过门槛。
  5. 逻辑与引用检查后必须读取 references/tier2_semantic_review.md,由当前宿主 AI 直接执行 Tier2 语义检查。由宿主 AI 自主规划术语、论证维度和可读性,输出“位置/障碍/影响/保真改法”。检查长句层级、指代、缩写/新概念、抽象名词、衔接…
  6. 可读性改写须保留已核验事实、对象、限定、问题/假设、引用、标签和 LaTeX 结构;无法保真保留原句并记录风险,不补事实/证据。

What it can do on your machine

Read from SKILL.md and the folder at commit b8b4142. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 1 file in scripts/, which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

NSFC Grant Rationale Writer loads about 945 tokens when it runs, and up to ~4.9k if it reads all its reference files. Until then it costs about 40 tokens; SKILL.md has 201 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~40
When it runs · the whole SKILL.md, loaded when a task matches
~945
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from huangwb8/ChineseResearchLaTeX at commit b8b4142, republished under its MIT licence (© huangwb8). 201 words, ~945 tokens.

Download SKILL.mdSave it as .claude/skills/nsfc-justification-writer/SKILL.md (or your agent's skills folder). This skill also uses 78 other files; get the full folder from GitHub.
name
nsfc-justification-writer
description
当用户要求写作、重构、审查或润色 NSFC/科研基金申请书的立项依据、研究意义、国内外现状、科学问题或科学假设时使用。以语义论证和可核验性为核心,支持任意文件名、标题命令和 LaTeX 结构;默认全自动完成诊断、成稿、校正、备份和写入,不向用户发起中途确认或审核问题。
metadata.author
Bensz Conan

科研立项依据写作器

流程

输入

按用户请求和配置文件提供必要输入;缺失信息应明确列出并停止依赖该输入的步骤。

执行步骤
  • 本 skill 设计缺陷导致 bug 时,按 bensz-collect-bugs 记录到 ~/.bensz-skills/bugs/,不要改用户本地已安装 skill;可用 workaround 时先记录再继续。
  • 仅用户明确要求公开上报时,才用本地 gh 通过该 skill 上传新增 bug 到 huangwb8/bensz-bugs;不 pull/clone 整仓库。
  1. 先写领域价值/必要性与已有证据,再将瓶颈转成待解释的关系、机制、性质或边界。
  2. 科学问题用疑问句追问认知缺口;假设用可证伪陈述句预测结果。研究目标、路线、验证终点不冒充问题/假设。
  3. 假设前以事实、文献和缺口为主;干预、比较、终点、路线放在假设后。外部文献的方法描述可作证据,不等于本项目方案前置。
  4. 切入点/贡献无需固定标题或段数;理论、混合、工程导向只改变措辞与证据重心,不是通过门槛。
  5. 逻辑与引用检查后必须读取 references/tier2_semantic_review.md,由当前宿主 AI 直接执行 Tier2 语义检查。由宿主 AI 自主规划术语、论证维度和可读性,输出“位置/障碍/影响/保真改法”。检查长句层级、指代、缩写/新概念、抽象名词、衔接和无效修饰;不科普化或因必要限定强改。
  6. 可读性改写须保留已核验事实、对象、限定、问题/假设、引用、标签和 LaTeX 结构;无法保真保留原句并记录风险,不补事实/证据。

措辞边界:吹牛式、绝对化或无依据夸大由宿主 AI 按 references/boastful_expression_guidelines.md 语义复核并归入逻辑/表达建议;Python 只做路径、结构命令、引用 key 等确定性检查,不用固定词表判定。

  1. 读取目标文件和上下文;目标缺失时自动追踪 main.tex,按确定性优先级选择候选,必要时创建白名单内的默认正文文件。
  2. 检查问题—假设—证据—研究内容闭环及引用 key;随后必做 Tier2,结果分为“事实问题、逻辑问题、表达建议”。
  3. 完成检查后自动复核专业可读性;polish 先在内部锁定事实/逻辑/引用,再做受约束精修,不向用户提问。
  4. 自动生成完整正文或 unified diff;发现结构命令变化时自动回退为正文-only 提案并重试。
  5. 通过白名单、引用和质量检查后自动写入;写入前备份,写入后展示 diff、假设与回滚入口。
  6. 字数与引用检查可独立运行;流程不绑定固定标题或维度清单。
  • scripts/run.py preview:自动生成 diff 并按 guardrails.output_mode 写入;需要只读检查时显式加 --dry-run。
  • refs、wordcount:独立引用/字数工具;术语、论证维度和可读性由 Tier2 宿主 AI 规划。
  • diagnose:仅执行 Tier1 确定性诊断,并提示宿主 AI 读取 references/tier2_semantic_review.md;不主动调用模型、不生成 Tier2 fallback 结论。
  • coach、review:主流程必须包含由宿主 AI 直接执行的 Tier2 语义检查。吹牛式/绝对化风险按参考准则由 AI 判断,脚本不做固定短语命中。
  • apply-section:legacy 标题替换写入;新流程优先使用完整正文提案自动写入,不依赖 \\subsubsection。
  • diff、rollback:查看或回滚自动写入变更;回滚命令仍需显式 --yes 以避免误操作。

按任务先阅读 references/tier2_semantic_review.md,再按其中需要阅读 references/scientific_question_guidelines.md、references/scientific_hypothesis_guidelines.md、references/professional_readability_guidelines.md 和 references/boastful_expression_guidelines.md。用户流程、脚本用法和架构边界见 README.md;检查维度由宿主 AI 自主规划,不依赖固定术语矩阵或四维度门槛。

输出

默认输出:

text
目标文件(自动选择及依据):...
修改范围:仅白名单正文
主要问题:...
建议正文或 unified diff:...
引用与事实守护:...
未修改:标题、结构命令、配置和样式文件

可读性发现归入“表达建议”,不当作事实错误;Tier2 是完整诊断必选项,由当前宿主 AI 直接完成,不经过 Python responder 或 fallback。

无法明确正文边界时按最小安全正文范围继续;若材料不足,采用保守假设并在“假设与待核验项”中记录,不暂停等待用户输入。始终执行白名单、备份、缺失 bibkey 检查并报告实际变更。

输出管理

本 Skill 的新任务中间文件统一写入 ./.bensz-api/task-{yyyymmdd-hhmm}-{简短描述}/{skill名}/input|output|log/。同一任务复用一个任务根目录;多 Skill 协作才创建 shared/。正式交付物不写入该目录,历史隐藏目录只允许显式兼容读取、迁移或清理。

校验

完成后执行 Skill 已有的静态检查、脚本验证或人工复核,并记录通过标准。

失败与恢复

保留错误证据和已完成产物;仅在输入、环境或外部依赖恢复后从最近的失败步骤重试。

约束

  • 自动理解目标文件、段落和目标,直接生成成稿并写入安全目标;同时输出 unified diff、备份路径和假设记录。
  • 只改自动确定且通过白名单的正文;保留标题、章节/环境命令、标签、引用命令和配置命令。
  • 只改白名单正文文件,不改 main.tex、配置文件、.cls、.sty。目标不唯一时按“已配置目标 → 文件名含立项依据/justification → 字典序首项”自动选择并记录选择依据。
  • 目标可来自用户或配置,不假定固定文件名。
  • 引用 key 必须能在项目 .bib 中核验;不新增无法核验的引用、结果或绝对化结论。
公共硬约束
  • 任务需要落盘时,使用唯一的 ./.bensz-api/task-{yyyymmdd-hhmm}-{简短描述}/ 根目录;共享材料放入 shared/,Skill 专属材料放入该 Skill 的 input/、output/、log/。
  • 正式交付物、源代码和正式计划按项目约定保存,不写入任务工作区;未经授权不覆盖、删除、迁移或远程写入。
  • 项目维护变更检查 BAC 可用性并记录需求、AI 产出、工具结果、文件改动和验证摘要;BAC 只做过程审计,不替代署名、责任或合规判断。
  • 不记录 API Key、访问令牌、密码、Cookie、环境/凭据文件、私有 Prompt、身份信息、本地用户名、主机名或不必要的大体积原始数据。
  • 文件路径必须规范化并限制在授权项目范围内;外部 URL、子进程和网络访问遵循最小权限,防止路径遍历、SSRF 和命令注入。
  • Skill 版本唯一记录在自身 config.yaml:skill_info.version;公开 API、协议、目录或配置变更同步文档与 CHANGELOG.md。
  • 仅将 Skill 或 Bensz 基础设施本身的设计缺陷交给 bensz-collect-bugs;先脱敏写入 ~/.bensz-skills/bugs/,当前任务不中断,只有用户明确要求才公开上报,禁止直接修改用户已安装的 Skill 源码。
<!-- End of canonical common constraints. -->
Skill 专属约束

不得超出本 Skill description 和上方流程所声明的范围;不将未验证的信息伪装成确定结论。

© huangwb8, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 78 other files (scripts, references, assets) in skills/nsfc-justification-writer of huangwb8/ChineseResearchLaTeX.

  • SKILL.md
  • CHANGELOG.md
  • README.md
  • assets/README.md
  • assets/examples/README.md
  • assets/examples/biology/example_001.metadata.yaml
  • assets/examples/biology/example_001.tex
  • assets/examples/chemistry/example_001.metadata.yaml
  • assets/examples/chemistry/example_001.tex
  • assets/examples/cs/example_001.metadata.yaml
  • assets/examples/cs/example_001.tex
  • assets/examples/engineering/example_001.metadata.yaml
  • assets/examples/engineering/example_001.tex
  • assets/examples/materials/example_001.metadata.yaml
  • … and 65 more

Open the folder on GitHubat commit b8b4142

Compare with similar skills

NSFC Grant Rationale Writer next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

NSFC Grant Rationale Writer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
NSFC Grant Rationale Writer this skillhuangwb8/ChineseResearchLaTeX2.9k—~945Automated safety check: PassMIT
Academic Paper Writing PipelineImbad0202/academic-research-skills51k—~16kAutomated safety check: PassCustom licence
Scientific Venue Templatesdavila7/claude-code-templates33k9 repos~5.1kAutomated safety check: NotesMIT
Econ Writehanlulong/econ-writing-skill6512 repos~14kAutomated safety check: PassMIT
Paper OrchestraAr9av/PaperOrchestra6791 repos~3.5kAutomated safety check: PassCustom licence
Backward Traceabilitylingzhi227/agent-research-skills390—~802Automated safety check: PassNone

Similar skills

  • Academic Paper Writing Pipeline

    Imbad0202/academic-research-skills

    Runs a 12-agent pipeline that plans, drafts, cites, reviews and formats academic papers, with modes for revision, rebuttals, abstracts and citation checks.

    51k GitHub stars~16k tokensUpdated today
    Research & ScienceAuto-check passed
  • Scientific Venue Templates

    davila7/claude-code-templates

    Supplies LaTeX templates and formatting rules for journals, conferences, posters, and grant proposals, then can check a draft against them.

    33k GitHub starsUsed in 9 repos~5.1k tokens
    Research & ScienceAuto-check: notes
  • Econ Write

    hanlulong/econ-writing-skill

    Expert economics paper writing assistant synthesizing advice from 50+ top guides by Cochrane, McCloskey, Shapiro, Head, Bellemare, Goldin, Glaeser, Kremer, and other leading economists.

    651 GitHub starsUsed in 2 repos~14k tokens
    Research & ScienceAuto-check passed
  • Paper Orchestra

    Ar9av/PaperOrchestra

    Orchestrate the full PaperOrchestra (Song et al., 2026, arXiv:2604.05018) five-agent pipeline to turn unstructured research materials (idea, experimental log, LaTeX template, conference guidelines…

    679 GitHub starsUsed in 1 repo~3.5k tokens
    Research & ScienceAuto-check passed
  • Backward Traceability

    lingzhi227/agent-research-skills

    Makes each number in a LaTeX paper link back to the code line that produced it, using hypertarget and hyperlink tags and compile-time `\num` formulas.

    390 GitHub stars~802 tokensUpdated 7 mo ago
    Research & ScienceAuto-check passed
  • Academic Venue Templates

    foryourhealth111-pixel/Vibe-Skills

    Supplies venue-specific LaTeX templates and formatting rules for journals, conferences and posters, and checks a manuscript against page limits and submission requirements.

    3.6k GitHub stars~3.9k tokensUpdated 1 mo ago
    Research & ScienceAuto-check: notes

More from huangwb8/ChineseResearchLaTeX

All 27 skills in this repo
  • LaTeX Example Content Generator

    huangwb8/ChineseResearchLaTeX

    Fills an existing LaTeX project with sample sections, tables and figure narratives, protecting the template structure and previewing before writing.

    2.9k GitHub starsUsed in 1 repo~744 tokens
    Auto-check passed
  • Journal Selector for Manuscripts

    huangwb8/ChineseResearchLaTeX

    Recommends journals for a manuscript by filtering a bundled impact-factor catalog, verifying scope and quality online, and writing a ranked Markdown report.

    2.9k GitHub stars~1.7k tokensUpdated 7 days ago
    Auto-check passed
  • NSFC Abstract Writer

    huangwb8/ChineseResearchLaTeX

    Writes Chinese and English abstracts for NSFC grant applications, with a recommended title and five alternatives, within set character limits.

    2.9k GitHub starsUsed in 1 repo~1.3k tokens
    Auto-check passed
  • NSFC Budget Justification Writer

    huangwb8/ChineseResearchLaTeX

    Writes a submission-ready NSFC budget justification as a LaTeX project and renders budget.pdf from your grant proposal text and supporting materials.

    2.9k GitHub starsUsed in 1 repo~1.4k tokens
    Auto-check passed
  • NSFC Application Code Recommender

    huangwb8/ChineseResearchLaTeX

    Recommends five pairs of NSFC application codes, primary and secondary, from a grant proposal's text and writes the reasons to a Markdown report.

    2.9k GitHub starsUsed in 1 repo~1.1k tokens
    Auto-check passed
  • NSFC Proposal Length Aligner

    huangwb8/ChineseResearchLaTeX

    Checks a Chinese NSFC grant proposal against section length budgets, reports where it runs short or long, and guides meaning-preserving expansion or trimming.

    2.9k GitHub starsUsed in 1 repo~1.3k tokens
    Auto-check passed

Works with

Questions about NSFC Grant Rationale Writer

What does NSFC Grant Rationale Writer do?

Writes, restructures, reviews and polishes the rationale section of NSFC research grant applications in LaTeX, with backups and a diff before every write. The target is the rationale section of a Chinese NSFC funding proposal: research significance, the state of the field, the scientific question and the hypothesis.tex` when none is given), checks that the problem, hypothesis, evidence and research content hang together and that citation keys exist, then runs a semantic review that sorts findings into factual problems, logic problems and expression suggestions.

When should I use NSFC Grant Rationale Writer?

NSFC Grant Rationale Writer fits situations like: drafting the research rationale section of an NSFC application in LaTeX; checking that a proposal's question, hypothesis and evidence line up; polishing dense sentences in a rationale section without changing facts or citations; counting words or auditing citation keys in a proposal section.

How do I install NSFC Grant Rationale Writer in Claude Code?

Run `npx skills add huangwb8/ChineseResearchLaTeX --skill nsfc-justification-writer -a claude-code`. Or copy the skill folder (skills/nsfc-justification-writer in huangwb8/ChineseResearchLaTeX) into .claude/skills/nsfc-justification-writer in your project. Claude Code loads it when a task matches its description.

How do I install NSFC Grant Rationale Writer in Codex?

Run `npx skills add huangwb8/ChineseResearchLaTeX --skill nsfc-justification-writer -a codex`. Or copy the skill folder (skills/nsfc-justification-writer in huangwb8/ChineseResearchLaTeX) into .agents/skills/nsfc-justification-writer in your project. Codex loads it when a task matches its description.

Can I use NSFC Grant Rationale Writer in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add huangwb8/ChineseResearchLaTeX --skill nsfc-justification-writer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/nsfc-justification-writer, .gemini/skills/nsfc-justification-writer, .github/skills/nsfc-justification-writer and .opencode/skills/nsfc-justification-writer in your project.

What does NSFC Grant Rationale Writer need to run?

SKILL.md names no scripts, command-line tools or credentials: NSFC Grant Rationale Writer is instructions for the agent only. Our summary lists: Python, to run `scripts/run.py`; A LaTeX grant project with a main file such as `main.tex`.

Does NSFC Grant Rationale Writer access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is NSFC Grant Rationale Writer safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does NSFC Grant Rationale Writer use?

NSFC Grant Rationale Writer is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does NSFC Grant Rationale Writer use?

About 945 tokens (SKILL.md is roughly 3.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4k tokens, read only when the agent opens those files.

What are the alternatives to NSFC Grant Rationale Writer?

Skills that share tags, products or a category with NSFC Grant Rationale Writer: Academic Paper Writing Pipeline (Imbad0202/academic-research-skills, 51k stars), Scientific Venue Templates (davila7/claude-code-templates, 33k stars), Econ Write (hanlulong/econ-writing-skill, 651 stars) and Paper Orchestra (Ar9av/PaperOrchestra, 679 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains NSFC Grant Rationale Writer?

huangwb8 (a GitHub user) maintains it in huangwb8/ChineseResearchLaTeX, which has 2,880 GitHub stars. The repository holds 27 skills in this directory. The repository was last updated on October 4, 2026.

Source: huangwb8/ChineseResearchLaTeX on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.