doc-research

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Chinese

Perform prioritized documentation research and return only relevant results.

执行优先级排序的文档检索，仅返回相关结果。

Priority	Tool	Use Case
Special	`mcp__plugin_totto2727_cloudflare-docs__search_cloudflare_documentation`	Cloudflare product queries only
Primary	`c7` skill (via Skill tool)	General library/framework official docs
Fallback	`WebSearch`	When above sources are insufficient
Fallback	`mcp__plugin_totto2727_cloudflare-browser__get_url_markdown`	Extract clean markdown from URLs found via WebSearch

优先级	工具	使用场景
特殊级	`mcp__plugin_totto2727_cloudflare-docs__search_cloudflare_documentation`	仅适用于Cloudflare产品相关查询
主选级	`c7` skill（通过Skill工具调用）	通用库/框架的官方文档查询
备选级	`WebSearch`	当前述来源无法提供足够信息时使用
备选级	`mcp__plugin_totto2727_cloudflare-browser__get_url_markdown`	从WebSearch找到的URL中提取干净的Markdown内容

Determine query type
- Cloudflare-related → Use
```
search_cloudflare_documentation
```
- General library/framework → Invoke
```
c7
```
  skill via Skill tool
Evaluate results
- Sufficient information found → Return results
- Insufficient → Proceed to step 3
Fallback: WebSearch
- Search with keywords targeting official sources (official docs, official blogs)
- Prefer results from official documentation sites
Deep content extraction (if needed)
- Use
```
get_url_markdown
```
  to extract clean markdown from promising URLs
- Only fetch URLs that are likely to contain the needed information

确定查询类型
- 与Cloudflare相关 → 使用
```
search_cloudflare_documentation
```
- 通用库/框架 → 通过Skill工具调用
```
c7
```
  skill
评估结果
- 已找到足够信息 → 返回结果
- 信息不足 → 进入步骤3
备选方案：WebSearch
- 使用针对官方来源的关键词进行搜索（官方文档、官方博客）
- 优先选择官方文档站点的结果
深度内容提取（如有需要）
- 使用
```
get_url_markdown
```
  从有价值的URL中提取干净的Markdown内容
- 仅获取可能包含所需信息的URL

Always start with the highest-priority source for the query type
Prefer official documentation over third-party content
Return concise, relevant results only — do not include excessive raw output
When using WebSearch, construct queries that target official sources
When multiple results are found, summarize the key information rather than dumping raw content