본문으로 건너뛰기
cccskills
무료GitHub에 공개

nsfc-grant-writer

Turn a scientific hypothesis into a Chinese application for the NSFC General Program, using its 2026 template. The workflow searches PubMed for real references, drafts the specified sections, and exports the result as Markdown and Word documents.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

원문원문 설명 보기

Generate a complete NSFC (National Natural Science Foundation of China) General Program grant application in Chinese, following the 2026 official template. Use when the user provides a scientific hypothesis and wants to generate a full grant application document, including abs...

연동 도구Word

실행 점수 미검증

설치 방법 보기

이럴 때 유용합니다

  • Drafting an NSFC General Program application from a hypothesis
  • Finding real PubMed references for a grant proposal
  • Preparing a Chinese application in the 2026 template
  • Exporting a grant draft to Word for completion

포함된 파일(7)

  • SKILL.md5.9 KB
  • eval_report_nsfc-grant-writer_result.json8.5 KB
  • POLISH_CHANGELOG.md658 B
  • references/pubmed-guide.md3.9 KB
  • references/template.md4.5 KB
  • references/writing-guide.md9.1 KB
  • scripts/generate_docx.py8.0 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Source: https://github.com/aipoch/medical-research-skills

NSFC Grant Writer

Generate a complete NSFC General Program grant application from a scientific hypothesis and export it as a Word document.

Workflow

Step 1: Parse hypothesis → extract key terms
Step 2: PubMed search → retrieve real references
Step 3: Generate full Markdown document (all sections)
Step 4: Save Markdown → call generate_docx.py → output .docx

Step 1: Parse the Hypothesis

Extract from the hypothesis:

  • Disease/condition
  • Key molecules (proteins, genes, miRNAs, etc.)
  • Biological process or pathway
  • Proposed mechanism (e.g., promotes, inhibits, regulates via)

Also generate the project title (≤30 Chinese characters):

  • Format: Mechanism of {key molecules}/{axis}/{pathway} regulating {biological process} or similar
  • Must reflect the core molecules, mechanism, and disease from the hypothesis
  • Example: Mechanism of KDM5D/NEDD4/Cx43/ATP axis in astrocytes regulating microglial activation and inflammatory response in NP
  • This title is used as the # [Title] heading at the top of the document

Do NOT ask the user for additional information unless they have provided it. Proceed with only the hypothesis.

Step 2: PubMed Literature Search

MANDATORY: All references must be retrieved via real PubMed API calls. Never fabricate references.

Read references/pubmed-guide.md for the full search protocol.

Quick reference:

  • Search: https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term={QUERY}&retmax=10&sort=relevance&retmode=json
  • Fetch summaries: https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&id={PMIDs}&retmode=json
  • Run 3-5 queries covering: disease epidemiology, key molecule expression, pathway/mechanism, therapeutic relevance
  • Select 15-25 total references; format in APA

Step 3: Generate the Full Document

Read references/writing-guide.md for detailed section-by-section requirements.

Generate the complete document as a single Markdown string following the structure in references/template.md.

Section generation rules:

SectionAction
AbstractGenerate Chinese + English abstract + keywords
Section I: Research RationaleGenerate full content: 5 numbered subsections with #### subheadings + hypothesis statement + research directions + reference list
Section II: Research ContentGenerate: 3-4 research modules + innovation points + 4-year plan
Section III-1.1: Preliminary Research FoundationGenerate: expand on preliminary data from Section I
Section III-1.2: Feasibility AnalysisGenerate: 4-dimension feasibility analysis
Section III (2-4) and Section IV (all subsections)Leave as headings only — do NOT generate content

Citation rules:

  • Inline: [N] or [N][M] (max 2 per location), placed before punctuation
  • Reference list at end of Section I only, APA format
  • Sections II and III reuse the same reference numbers if needed

Consistency rule: Every preliminary data item mentioned in Section I (Figure 1, Figure 2, Table 1, etc.) MUST appear again in Section III 1.1, expanded with methods and results detail.

Step 4: Create Output Folder and Export

Output folder naming

Extract 2-3 core keywords from the hypothesis (Chinese preferred, use underscores). Format: {keyword1}_{keyword2}_{YYYYMMDD_HHMMSS} Example: EGFR_LungCancer_20260527_143022

Create this folder in the current working directory (or user's specified path).

Save Markdown

Save the full generated document as {folder_name}.md inside the output folder (same stem as the folder name).

Generate Word document

Run the script:

python scripts/generate_docx.py {output_folder}/{folder_name}.md {output_folder}

This produces {output_folder}/{folder_name}.docx.

Dependency check

If python-docx is not installed, inform the user:

pip install python-docx

Then re-run the script.

Final output to user

After successful generation, report:

  • Output folder path
  • Files created: {folder_name}.md and {folder_name}.docx
  • Note: Sections requiring personal information (working conditions, ongoing projects, completed projects, Part IV) are left as headings for the applicant to complete
  • Note: The mechanism diagram placeholder in Section I requires replacement with an actual figure

Resources

  • references/template.md — 2026 official template structure
  • references/writing-guide.md — detailed writing requirements per section
  • references/pubmed-guide.md — PubMed API search protocol and reference formatting
  • scripts/generate_docx.py — Markdown to Word converter (requires python-docx)

Error Handling

  • If required inputs are missing, state exactly which fields are missing and request only the minimum additional information.
  • If the task goes outside the documented scope, stop instead of guessing or silently widening the assignment.
  • If execution fails, report the failure point, summarize what can still be completed safely, and provide a manual fallback.
  • Do not fabricate files, citations, data, search results, or execution outcomes.

Input Validation

This skill accepts requests that match the documented purpose of nsfc-grant-writer and include enough context to complete the workflow safely.

Do not continue the workflow when the request is out of scope, missing a critical input, or would require unsupported assumptions. Instead respond:

nsfc-grant-writer only handles its documented workflow. Please provide the missing required inputs or switch to a more suitable skill.

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Condense an academic paper into a structured abstract of up to 250 words, organized by background, objective, methods, results, and conclusion. It aims to retain key quantitative findings and supports papers from several research fields; figures and charts are not interpreted.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Creating a concise abstract from a research paper
  • Summarizing empirical studies while retaining key statistics
  • Preparing structured summaries across multiple papers
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

Trim English academic abstracts to a target word count while preserving key findings, statistics, and the background-to-conclusion structure. Choose conservative, balanced, or aggressive compression, check word counts, and review the edited text for readability and accuracy.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Cutting an abstract to meet a journal word limit
  • Preparing a conference abstract with a strict word count
  • Checking an abstract's word count before submission
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

Refines long medical drafts, reports, or literature notes into formal, single-paragraph Chinese and English abstracts in an SCI journal style. It can place both abstracts in a Markdown report while keeping claims grounded in the source text.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Condensing a medical study report into bilingual abstracts
  • Preparing unstructured abstracts from literature notes
  • Creating a Markdown report with Chinese and English abstracts
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

Creates three to five concise English Highlights for journal submission from manuscript text or supported files. It adapts bullets to the article type, keeps each within 85 characters, and grounds every claim in the supplied source without inventing results.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Draft journal Highlights from an abstract or manuscript
  • Summarize a meta-analysis or review for submission
  • Extract key findings from a supported PDF or DOCX
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

Design comparative drug safety studies using FAERS, the FDA’s adverse event reporting database, within one defined safety area. The plan compares drugs or classes, lays out workload options and analysis steps, and distinguishes detected signals from conclusions the data cannot support.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Comparing drug classes within one adverse event area
  • Planning a FAERS study with active comparators
  • Adding declared subgroup or onset characterization
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

Predicts absorption, distribution, metabolism, and excretion properties for small molecules from their structures. Use the results to compare compounds, screen libraries, and review drug-likeness scores in structured outputs. Predictions support prioritization and need experimental validation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Compare predicted ADME profiles for small-molecule candidates
  • Screen a compound library for drug-likeness
  • Rank candidates against target property ranges
aipoch/medical-research-skills2천2026년 9월 17일 업데이트

aipoch의 스킬 모두 보기

이 스킬의 문제 신고