literature-review-writer

Top-tier journal literature review writing tool

Core Positioning

Write a Literature Review section that meets top-tier journal standards for the research topic provided by the user. Unlike the encyclopedic coverage of a doctoral dissertation literature review, a top-tier journal literature review is characterized by conciseness, argument-driven writing, and strategic selection of literature.

Workflow

Writing a literature review involves the following steps:

1. Parse the research topic and break down the core constructs and variable relationships
2. Retrieve real literature (WebSearch)
3. Verify the authenticity of the literature (refer to references/verification-guide.md)
4. Determine the review structure (refer to references/structure-patterns.md)
5. Write the main body of the review (refer to references/writing-techniques.md)
6. Output the complete literature review and reference list (dialogue output + generate Word document)

Step 1: Parse the Research Topic

Identify from the research topic provided by the user:

• Core independent variable (IV)anddependent variable (DV)
• Possible mediator (Mediator)andmoderator (Moderator)
• The disciplinary field and subfield in which the research is situated
• Potentially applicable theoretical frameworks

Break down the topic into 3-5 thematic modules that need to be reviewed.

Step 2: Retrieve Real Literature

Use WebSearch to systematically search the literature for each thematic module. Search strategy:

Required searches：

• For each core construct:"[构念]" site:scholar.google.com or "[构念]" review meta-analysis
• Core variable relationships:"[IV]" "[DV]" empirical study
• Theoretical foundations:"[理论名]" "[研究领域]"
• Top-tier journal targeted searches:"[核心概念]" site:journals.sagepub.com OR site:onlinelibrary.wiley.com OR site:sciencedirect.com

Literature screening priorities：

1. Empirical research in top-tier journals from the past 5 years (most important)
2. Seminal works in the field (regardless of year)
3. Recent meta-analyses or systematic reviews
4. Highly cited classic theoretical literature

Target number of references: Find 5-10 high-quality references for each thematic module, with the total controlled at 25-50.

Step 3: Verify the Authenticity of the Literature

This is a mandatory step and cannot be skipped. It must be completed before writing the main body of the review.

Call literature-verifier skill, and submit all the references to be cited retrieved in Step 2 for verification. This skill can:

• Cross-check metadata such as authors, titles, journals, and years
• Detect fake literature generated by AI hallucinations
• Cover Chinese and English databases (Google Scholar, CNKI, Wanfang, VIP, etc.)
• Verify the validity of DOIs

Verification rules:

• All references to be cited must pass literature-verifier verification before they can be used
• References that fail verification must be removed immediately and must never be cited
• If a thematic module has insufficient literature after removal, return to Step 2 to supplement the search and verify again

Step 4: Determine the Review Structure

Read references/structure-patterns.md to select an appropriate structure.

By default, use a funnel structure(unless the research topic clearly suits another structure):

第一层：宏观领域定位（1段）
第二层：核心构念的文献脉络（2-4段，按主题组织）
第三层：变量间关系的研究现状（1-3段）
第四层：研究空白与本文定位（1段）

Step 5: Write the review body

Read references/writing-techniques.md for writing technique guidance.

Writing Iron Rules：

1. Thematic synthesis, not article-by-article listing: Each paragraph centers on one argument, synthesizing multiple studies for support
2. Each paragraph must have a clear argumentative function: Do not write paragraphs without a purpose
3. Connect paragraphs with logical transition sentences: Reflect progression, contrast, or focal relationships
4. Embed critical evaluation in the argumentation: Point out limitations in methodology, theory, or empirical coverage
5. The literature thread is clearly visible: Present the research evolution trend (pioneering → development → frontier → gap)
6. Conclude with the research gap: Clearly point out the gap and naturally lead to the necessity of this study

Citation format: Use APA 7th edition by default. If the user specifies another format, follow it.

Language：

• If the user provides the topic in Chinese, write the review in English (top-journal standard), unless the user explicitly requests Chinese
• Write academically and concisely, avoiding colloquial expressions

Step 6: Output format

6a: Conversation output

Directly output the following in the conversation:

## Literature Review

[完整的文献综述正文，含嵌入式引用]

## References

[按APA 7th或用户指定格式列出所有引用文献的完整信息]

Format of each reference: Author, A. A., & Author, B. B. (Year). Title of article. Journal Name, Volume(Issue), Pages. https://doi.org/xxx

6b: Generate Word document

After the conversation output is complete,mustalso generate a Word document:

1. Write the review body and reference list into a temporary markdown file
2. Call scripts/generate_docx.py to generate the .docx file

Operation steps：

# 1. 将综述内容写入临时 md 文件
cat > /tmp/lit_review_temp.md << 'EOF'
[此处粘贴完整的综述 markdown 内容，包含 ## Literature Review 和 ## References]
EOF

# 2. 生成 Word 文档（输出到用户桌面）
python3 [skill_base_dir]/scripts/generate_docx.py \
  --input /tmp/lit_review_temp.md \
  --output ~/Desktop/Literature_Review.docx \
  --title "[论文标题]"

Word document format specifications：

• Body: Times New Roman 12pt, double-spaced, first-line indent 0.75cm
• Chinese font fallback: SimSun (body), SimHei (headings)
• References: hanging indent 1.27cm
• Margins: 1 inch on all sides

File naming: By default, save to ~/Desktop/Literature_Review.docx. If the user specifies a path, follow it.

Script description：scripts/generate_docx.py accepts the following parameters:

• --input: Input markdown file path (required)
• --output: Output .docx file path (required)
• --title: Document title, displayed centered and bold at the top of the document (optional)

Quality checklist

After completing the writing, self-check against the following checklist:

• Is thematic organization used rather than article-by-article listing?
• Does each paragraph have a clear topic sentence?
• Are there logical transitions between paragraphs?
• Is the evolutionary thread of the literature presented?
• Are consensus and disagreement synthesized?
• Is critical evaluation included?
• Is the research gap clearly identified?
• Does the gap naturally point to the necessity of this study?
• Have all cited references been verified for authenticity?
• Is the citation density reasonable (5-12 per paragraph)?
• Is the reference list complete and consistently formatted?
• Has the Word document been generated and the file path communicated to the user?