A structured web research workflow, with the core goal: persist research results in real time to prevent work from being lost due to session truncation.
When to Use
- Preliminary research for writing an article
- Learn about new products, new technologies, new releases
- Gather competitor information or industry trends
- Any information search task that requires multiple WebSearch calls
Execution Workflow
Step 1: Immediately create a research file
- Before starting the search, create the file first
- Path:
_knowledge_base/research-<主题>-<YYYYMMDD>.md - Initial content includes: research goals, key questions, expected output
# [主题] 调研笔记
调研日期：YYYY-MM-DD
调研目标：[一句话说明]
## 关键问题
1. [问题1]
2. [问题2]
3. [问题3]
## 发现
（调研中逐步填充）
## 来源列表
（每次搜索后追加）
Step 2: Search and save incrementally
- After each WebSearch, immediately append the findings to the file
- Attach the source URL and date to each finding
- Follow the information source priority (see SHARED-RULES.md)
Step 3: Stage summary
- After every 3 searches, save a "stage summary" in the file
- Format:
### 阶段摘要 (第N轮)+ current key findings
Step 4: Final brief
At the end of the research, organize the file into a structured brief:
## 调研结论
### 关键事实
1. [事实1]（来源：URL）
2. [事实2]（来源：URL）
### 来源列表
| 来源 | URL | 发布日期 | 可信度 |
|------|-----|---------|--------|
| ... | ... | ... | 高/中/低 |
### 待确认问题
- [还需要进一步验证的点]
### 写作建议
- [基于调研结果，对后续写作的建议]
Key Principles
- Create the file before searching: Ensure that the first search result is saved
- Save incrementally, do not wait until the end: Append immediately after each search
- Separate research from writing: This Skill only does research, and does not start writing a draft
- Mark credibility: Distinguish first-hand information (official) from second-hand information (media/community)
- Ignore outdated information sources: Zhihu/Baidu (before 2025), marketing advertorials
Relationship with Other Skills
- After research is complete, the user can trigger /topic generation to determine the writing direction
- The research file will serve as input material for subsequent writing
- If information found during research is suitable for long-term retention, save it to the corresponding _knowledge_base category directory
Output Location
- Research notes:
_knowledge_base/research-<主题>-<YYYYMMDD>.md - Long-term knowledge:
_knowledge_base/<分类>/<主题>-<YYYYMM>.md
Last updated: 2026-02-06
Produced by Huashu | AI Native Coder · Independent Developer WeChat public account "Huashu" | 300K+ followers | AI tools and efficiency improvement Representative works: Xiaomao Fill Light (AppStore paid chart Top1) · "One Book to Master DeepSeek"