mazdac/Szpitale-graph
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: enhance research workflow with user confirmations

Browse Source 
- Add AskUserQuestion for Step 1 framework confirmation
- Add TOC summary field selection in report generation
- Add detail_level hierarchy (brief -> moderate -> detailed)
- Add Chinese output requirement for JSON values
- Add long text formatting rules

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
jilinchen
2025-12-31 11:45:24 +08:00
co-authored by Claude Opus 4.5
parent 5eaf859274
commit 11a51a8d79
4 changed files with 276 additions and 251 deletions
Download Patch File Download Diff File Expand all files Collapse all files
skills/research/SKILL.md
+92 -89
View File
@@ -1,140 +1,143 @@
---
allowed-tools: Read, Write, Glob, WebSearch, Task, AskUserQuestion
description: Conduct preliminary research on a topic and generate a research outline. Use for academic research, benchmark research, technology selection, etc.
description: 对目标话题进行初步调研，生成调研outline。用于学术调研、benchmark调研、技术选型等场景。
---
# Research Skill - Preliminary Research
# Research Skill - 初步调研
## Trigger
## 触发方式
`/research <topic>`
## Workflow
## 执行流程
### Step 1: Generate Initial Framework from Model Knowledge
Based on the topic, use model's existing knowledge to generate:
- Main research objects/items list in this domain
- Suggested research field framework
### Step 1: 模型内部知识生成初步框架
基于topic，利用模型已有知识生成：
- 该领域的主要研究对象/items列表
- 建议的调研字段框架
Output {step1_output}.
输出{step1_output}，使用AskUserQuestion确认：
- items列表是否需要增减？
- 字段框架是否满足需求？
### Step 2: Web Search Supplement
Use AskUserQuestion to ask for time range (e.g., last 6 months, since 2024, unlimited).
### Step 2: Web Search补充
使用AskUserQuestion询问时间范围（如：最近6个月、2024年至今、不限）。
**Parameter Retrieval**:
- `{topic}`: User input research topic
- `{YYYY-MM-DD}`: Current date
- `{step1_output}`: Complete output from Step 1
- `{time_range}`: User specified time range
**参数获取**：
- `{topic}`: 用户输入的调研话题
- `{YYYY-MM-DD}`: 当前日期
- `{step1_output}`: Step 1生成的完整输出内容
- `{time_range}`: 用户指定的时间范围
**Hard Constraint**: The following prompt must be strictly reproduced, only replacing variables in {xxx}, do not modify structure or wording.
**硬约束**：以下prompt必须严格复述，仅替换{xxx}中的变量，禁止改写结构或措辞。
Launch 1 web-search-agent (background), **Prompt Template**:
启动1个web-search-agent（后台），**Prompt模板**：
```python
prompt = f"""## Task
Research topic: {topic}
Current date: {YYYY-MM-DD}
prompt = f"""## 任务
调研话题: {topic}
当前日期: {YYYY-MM-DD}
Based on the following initial framework, supplement latest items and recommended research fields.
基于以下初步框架，补充最新items和推荐调研字段。
## Existing Framework
## 已有框架
{step1_output}
## Goals
1. Verify if existing items are missing important objects
2. Supplement items based on missing objects
3. Continue searching for {topic} related items within {time_range} and supplement
4. Supplement new fields
## 目标
1. 验证已有items是否遗漏重要对象
2. 根据遗漏对象进行补充items
3. 继续搜索{topic}相关且{time_range}内的items并补充
4. 补充新fields
## Output Requirements
Return structured results directly (do not write files):
## 输出要求
直接返回结构化结果（不写文件）：
### Supplementary Items
- item_name: Brief explanation (why it should be added)
### 补充Items
- item_name: 简要说明（为什么应该加入）
...
### Recommended Supplementary Fields
- field_name: Field description (why this dimension is needed)
### 推荐补充字段
- field_name: 字段描述（为什么需要这个维度）
...
### Sources
- [Source1](url1)
- [Source2](url2)
### 信息来源
- [来源1](url1)
- [来源2](url2)
"""
```
**One-shot Example** (assuming researching AI Coding History):
**One-shot示例**（假设调研AI Coding发展史）：
```
## Task
Research topic: AI Coding History
Current date: 2025-12-30
## 任务
调研话题: AI Coding 发展史
当前日期: 2025-12-30
Based on the following initial framework, supplement latest items and recommended research fields.
基于以下初步框架，补充最新items和推荐调研字段。
## Existing Framework
### Items List
1. GitHub Copilot: Developed by Microsoft/GitHub, first mainstream AI coding assistant
2. Cursor: AI-first IDE, based on VSCode
## 已有框架
### Items列表
1. GitHub Copilot: Microsoft/GitHub开发，首个主流AI编程助手
2. Cursor: AI-first IDE，基于VSCode
...
### Field Framework
- Basic Info: name, release_date, company
- Technical Features: underlying_model, context_window
### 字段框架
- 基本信息: name, release_date, company
- 技术特性: underlying_model, context_window
...
## Goals
1. Verify if existing items are missing important objects
2. Supplement items based on missing objects
3. Continue searching for AI Coding History related items within since 2024 and supplement
4. Supplement new fields
## 目标
1. 验证已有items是否遗漏重要对象
2. 根据遗漏对象进行补充items
3. 继续搜索AI Coding 发展史相关且2024年至今内的items并补充
4. 补充新fields
## Output Requirements
Return structured results directly (do not write files):
## 输出要求
直接返回结构化结果（不写文件）：
### Supplementary Items
- item_name: Brief explanation (why it should be added)
### 补充Items
- item_name: 简要说明（为什么应该加入）
...
### Recommended Supplementary Fields
- field_name: Field description (why this dimension is needed)
### 推荐补充字段
- field_name: 字段描述（为什么需要这个维度）
...
### Sources
- [Source1](url1)
- [Source2](url2)
### 信息来源
- [来源1](url1)
- [来源2](url2)
```
### Step 3: Ask User for Existing Fields
Use AskUserQuestion to ask if user has existing field definition file, if so read and merge.
### Step 3: 询问用户已有字段
使用AskUserQuestion询问用户是否有已定义的字段文件，如有则读取并合并。
### Step 4: Generate Outline (Separate Files)
Merge {step1_output}, {step2_output} and user's existing fields, generate two files:
### Step 4: 生成Outline（分离文件）
合并{step1_output}、{step2_output}和用户已有字段，生成两个文件：
**outline.yaml** (items + config):
- topic: Research topic
- items: Research objects list
**outline.yaml**（items + 配置）：
- topic: 调研主题
- items: 调研对象列表
- execution:
- batch_size: Number of parallel agents (confirm with AskUserQuestion)
- items_per_agent: Items per agent (confirm with AskUserQuestion)
- output_dir: Results output directory (default: ./results)
- batch_size: 并行agent数量（需AskUserQuestion确认）
- items_per_agent: 每个agent调研项目数（需AskUserQuestion确认）
- output_dir: 结果输出目录（默认./results）
**fields.yaml** (field definitions):
- Field categories and definitions
- Each field's name, description, detail_level
- uncertain: Uncertain fields list (reserved field, auto-filled in deep phase)
**fields.yaml**（字段定义）：
- 字段分类和定义
- 每个字段的name、description、detail_level
- detail_level分层：极简 → 简要 → 详细
- uncertain: 不确定字段列表（保留字段，deep阶段自动填充）
### Step 5: Output and Confirm
- Create directory: `./{topic_slug}/`
- Save: `outline.yaml` and `fields.yaml`
- Show to user for confirmation
### Step 5: 输出并确认
- 创建目录: `./{topic_slug}/`
- 保存: `outline.yaml` 和 `fields.yaml`
- 展示给用户确认
## Output Path
## 输出路径
```
{current_working_directory}/{topic_slug}/
├── outline.yaml # items list + execution config
└── fields.yaml # field definitions
{当前工作目录}/{topic_slug}/
├── outline.yaml # items列表 + execution配置
└── fields.yaml # 字段定义
```
## Follow-up Commands
- `/research-add-items` - Supplement items
- `/research-add-fields` - Supplement fields
- `/research-deep` - Start deep research
## 后续命令
- `/research-add-items` - 补充items
- `/research-add-fields` - 补充字段
- `/research-deep` - 开始深度调研