Three steps, each step provides multiple options + rationales, and after the user selects, proceed to the next step.
Start: collect necessary information
Collect at startup (information already available in ta-research-AFP is automatically reused and not asked again):
- Research topic and research question
- Core findings: topic summary table, or a brief description of 3–5 main findings
- Theoretical positioning(A/B/C) and the main theories used (if any)
- Target journal type: C journal / SSCI
Step 1: Title generation
Generate 4 candidate titles, covering different writing angles, each with a complete analysis.
Four title angles：
A. Phenomenon-focused type
- Format reference:
[研究对象]的[核心经验/行为/过程]：基于[N]位[被访者]的主题分析 - Features: places the research object and core phenomenon first, with the method clearly visible
B. Theoretical dialogue type
- Format reference:
[理论概念]视角下[研究对象]的[机制/过程]研究 - Features: places the theoretical contribution first, suitable for theory-driven (positioning A) research
C. Question-oriented type
- Format reference:
[研究对象]如何[行动/应对/经历][研究现象]？——基于[N]位[被访者]的质性研究 - Features: directly presents the research question, so readers immediately understand the research intent
D. Conclusion-oriented type
- Format reference:
[核心发现/命名]：[研究对象]在[情境]中的[经验/机制] - Features: uses the core finding as the title focus, the most attractive, but the finding must be clear enough to be named
Output format：
━━ 候选标题 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
【选项 A】现象聚焦型
标题：[具体标题]
角度：前置[研究对象]和[核心现象]，方法标注清晰
期刊适配：C刊（审稿人易判断研究范围）/ SSCI 均可
优势：[说明]
风险：[说明]
【选项 B】理论对话型
标题：[具体标题]
角度：前置[理论概念]，信号该研究具有理论贡献
期刊适配：更适合 [C刊/SSCI]（理由：[说明]）
优势：[说明]
风险：[说明]
【选项 C】问题导向型
标题：[具体标题]
角度：直接提出研究问题，读者即刻理解研究意图
期刊适配：均可
优势：[说明]
风险：[说明]
【选项 D】结论导向型
标题：[具体标题]
角度：以[核心发现名称]为题眼，最具吸引力
期刊适配：更适合 [C刊/SSCI]（理由：[说明]）
优势：[说明]
风险：需要发现足够清晰，否则标题会显得空洞
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
请选择一个标题（A/B/C/D），或说明修改方向。
⚠️ Hard stop point: after waiting for the user to select a title, proceed to Step 2.
Step 2: Abstract generation
Based on the selected title, generate 3 abstract versions, each with a structural explanation and applicable scenarios.
Three abstract structures：
Version A: Standard sequential structure Order: research background → research question → method → core findings → theoretical contribution Features: the most common structure, with each element clear, making it easy for reviewers to evaluate item by item Length: C journal 250–350 characters; SSCI 150–200 words
Version B: Contribution-first structure Order: core contribution statement → how it was derived (method + data) → main findings → theoretical significance Features: the first sentence directly states the research contribution, highly attractive, suitable for research with a clear contribution Length: C journal 250–300 characters; SSCI 150–200 words
Version C: Finding-first structure Order: core finding → research background and question → method → theoretical contribution Features: begins with the most important finding, suitable for research whose findings are especially prominent and whose readers most want to know "what the conclusion is" Length: depends on journal requirements, same as Version A
Output format：
━━ 候选摘要 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
【版本 A】标准顺序结构（[字数]字）
结构：背景 → 问题 → 方法 → 发现 → 贡献
适配：C刊首选，SSCI 均可
---
[摘要正文]
---
【版本 B】贡献前置结构（[字数]字）
结构：贡献声明 → 如何得出 → 发现 → 意义
适配：SSCI 首选，研究贡献清晰时也适合 C刊
---
[摘要正文]
---
【版本 C】发现前置结构（[字数]字）
结构：发现 → 背景 → 方法 → 贡献
适配：发现特别突出的研究，C刊/SSCI 均可
---
[摘要正文]
---
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
请选择一个版本（A/B/C），或说明修改方向。
⚠️ Hard stop point: after waiting for the user to select an abstract, proceed to Step 3.
Step 3: Keyword generation
Based on the selected title and abstract, generate 2 sets of keyword options, each with 5–8 words, and each word with a rationale for inclusion.
Two keyword strategies：
Plan A: Topic search type
- Purpose: Enable readers within the research field to find this paper through search
- Focus: Core concepts of the research subject, research context terms, research phenomenon terms
- Suitable for: C-journal submissions, increasing the probability of being cited by researchers in the same field
Option B: Theory + Method type
- Purpose: Enable readers interested in methodology and theoretical dialogue to find this paper
- Focus: Theory names/core concepts, method names (thematic analysis/qualitative research), disciplinary field terms
- Suitable for: SSCI submissions, attracting methodology interlocutors and theoretical researchers
Output format：
━━ 候选关键词 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
【方案甲】议题检索型（[N]个词）
关键词：[词1] / [词2] / [词3] / [词4] / [词5] / ...
各词纳入理由：
- [词1]：[为什么纳入，检索价值是什么]
- [词2]：[为什么纳入]
- ...
【方案乙】理论+方法型（[N]个词）
关键词：[词1] / [词2] / [词3] / [词4] / [词5] / ...
各词纳入理由：
- [词1]：[为什么纳入，检索价值是什么]
- [词2]：[为什么纳入]
- ...
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
请选择一个方案（甲/乙），或说明自定义组合方式。
⚠️ Hard stop point: After waiting for the user to select keywords, proceed to save.
Output and save
After all three steps are completed, call the Write tool to save:
File naming:title_abstract_keywords.md
Save path: current project directory
File format:
# 标题
[选定的标题]
# 摘要
[选定的摘要正文]
# 关键词
[选定的关键词，以"/"分隔]
Inform the researcher:
"Title/Abstract/Keywords have been saved to
title_abstract_keywords.md. If in ta-research-AFP, please continue to the next checkpoint (full-text citation verification)."
Language
- Default Chinese
- SSCI journals: Ask whether an English version needs to be generated simultaneously (English counterpart of title + abstract + keywords)
- English counterpart terms of keywords should also be annotated when saving