This is a structured web research workflow that saves your findings incrementally so you don't lose work when the conversation cuts off. It enforces a specific pattern: create a markdown file in _knowledge_base before you start searching, append discoveries after each WebSearch call, write interim summaries every three searches, and compile a final brief with sourced facts and credibility ratings. The author clearly got burned by losing research mid-session and built guardrails around it. If you do multi-step research tasks where you're gathering information from the web and need that work preserved between context windows, this gives you a disciplined template. It's opinionated about file naming, markdown structure, and separating research from writing.

skills.sh installs
753
GitHub stars
2.2K
last updated
Aug 21, 2026

Install

$ npx -y skills add alchaincyf/huashu-skills --skill huashu-research --agent claude-code

Installs into .claude/skills of the current project.

Files

调研 Skill

结构化的网络调研流程,核心目标:调研成果实时持久化,防止会话截断丢失工作。

何时使用

  • 为写文章做前期调研
  • 了解新产品、新技术、新发布
  • 搜集竞品信息或行业动态
  • 任何需要多次 WebSearch 的信息搜索任务

执行流程

Step 1: 立即创建调研文件

  • 在开始搜索之前,先创建文件
  • 路径:_knowledge_base/research-<主题>-<YYYYMMDD>.md
  • 初始内容包含:调研目标、关键问题、预期输出
# [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: 搜索并增量保存

  • 每次 WebSearch 后,立即将发现追加到文件
  • 每条发现附上来源 URL 和日期
  • 遵循信息源优先级(见 SHARED-RULES.md)

Step 3: 阶段摘要

  • 每完成3次搜索,在文件中保存一次「阶段摘要」
  • 格式:### 阶段摘要 (第N轮) + 当前关键发现

Step 4: 最终简报

调研结束时,整理文件为结构化简报:

## 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]

关键原则

  • 先建文件再搜索:确保第一次搜索结果就被保存
  • 增量保存不等到最后:每次搜索后立即追加
  • 调研和写作分离:本 Skill 只做调研,不开始写草稿
  • 标注可信度:区分一手信息(官方)和二手信息(媒体/社区)
  • 忽略过时信息源:知乎/百度(2025年前)、营销软文

与其他 Skill 的关系

  • 调研完成后,用户可触发 /选题生成 来确定写作方向
  • 调研文件将作为后续写作的输入素材
  • 如果调研中发现的信息适合长期留存,保存到对应的 _knowledge_base 分类目录

输出位置

Sponsors

Details

Repository
github.com/alchaincyf/huashu-skills
Category
Data Science & ML
First seen
Jun 3, 2026

More from alchaincyf/huashu-skills

All 21
1ruvnet avatarresearch-synthesizeruvnet/rufloSynthesize research findings from memory into structured reports with evidence grading, contradiction resolution, and actionable recommendations752skills.sh installs
2alinaqi avatarposthog-analyticsalinaqi/claude-bootstrapposthog analytics736skills.sh installs
3alirezarezvani avataranalytics-trackingalirezarezvani/claude-skillsSet up, audit, and debug analytics tracking implementation — GA4, Google Tag Manager, event taxonomy, conversion tracking, and data quality. Use when building a731skills.sh installs
4davila7 avatarcontext7-auto-researchdavila7/claude-code-templatesAutomatically fetch latest library/framework documentation for Claude Code via Context7 API724skills.sh installs
5drshailesh88 avatarsocial-media-trends-researchdrshailesh88/integrated_content_osProgrammatic social media and marketing research using free tools: pytrends (Google Trends), yars (Reddit without API keys), and Perplexity MCP (Twitter/TikTok/720skills.sh installs
6weizhena avatarresearch-add-fieldsweizhena/deep-research-skillsAdd field definitions to existing research outline.708skills.sh installs