github-deep-research by bytedance/deer-flow
npx skills add https://github.com/bytedance/deer-flow --skill github-deep-research结合 GitHub API、web_search、web_fetch 进行多轮研究,生成全面的 Markdown 报告。
由宽泛到聚焦:从 GitHub API 开始,然后进行一般性查询,根据发现结果进行细化。
Round 1: GitHub API
Round 2: "{topic} overview"
Round 3: "{topic} architecture", "{topic} vs alternatives"
Round 4: "{topic} issues", "{topic} roadmap", "site:github.com {topic}"
来源优先级:
第 1 轮 - GitHub API 直接执行 scripts/github_api.py,无需使用 read_file():
python /path/to/skill/scripts/github_api.py <owner> <repo> summary
python /path/to/skill/scripts/github_api.py <owner> <repo> readme
python /path/to/skill/scripts/github_api.py <owner> <repo> tree
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
可用命令(github_api.py 的最后一个参数):
第 2 轮 - 探索发现(3-5 次 web_search)
第 3 轮 - 深度调查(5-10 次 web_search + web_fetch)
第 4 轮 - 深入挖掘
遵循 assets/report_template.md 中的模板:
在有助于理解的地方包含图表:
时间线(甘特图):
gantt
title Project Timeline
dateFormat YYYY-MM-DD
section Phase 1
Development :2025-01-01, 2025-03-01
section Phase 2
Launch :2025-03-01, 2025-04-01
架构(流程图):
flowchart TD
A[User] --> B[Coordinator]
B --> C[Planner]
C --> D[Research Team]
D --> E[Reporter]
比较(饼图/柱状图):
pie title Market Share
"Project A" : 45
"Project B" : 30
"Others" : 25
根据来源质量分配置信度:
| 置信度 | 标准 |
|---|---|
| 高(90%+) | 官方文档、GitHub 数据、多个相互佐证的来源 |
| 中(70-89%) | 单一可靠来源、近期文章 |
| 低(50-69%) | 社交媒体、未经证实的说法、过时信息 |
将报告保存为:research_{topic}_{YYYYMMDD}.md
每周安装次数
222
代码库
GitHub Stars
27.8K
首次出现
Feb 17, 2026
安全审计
安装于
opencode218
gemini-cli217
github-copilot217
codex216
kimi-cli215
amp215
Multi-round research combining GitHub API, web_search, web_fetch to produce comprehensive markdown reports.
Broad to Narrow : Start with GitHub API, then general queries, refine based on findings.
Round 1: GitHub API
Round 2: "{topic} overview"
Round 3: "{topic} architecture", "{topic} vs alternatives"
Round 4: "{topic} issues", "{topic} roadmap", "site:github.com {topic}"
Source Prioritization :
Round 1 - GitHub API Directly execute scripts/github_api.py without read_file():
python /path/to/skill/scripts/github_api.py <owner> <repo> summary
python /path/to/skill/scripts/github_api.py <owner> <repo> readme
python /path/to/skill/scripts/github_api.py <owner> <repo> tree
Available commands (the last argument ofgithub_api.py):
Round 2 - Discovery (3-5 web_search)
Round 3 - Deep Investigation (5-10 web_search + web_fetch)
Round 4 - Deep Dive
Follow template in assets/report_template.md:
Include diagrams where helpful:
Timeline (Gantt) :
gantt
title Project Timeline
dateFormat YYYY-MM-DD
section Phase 1
Development :2025-01-01, 2025-03-01
section Phase 2
Launch :2025-03-01, 2025-04-01
Architecture (Flowchart) :
flowchart TD
A[User] --> B[Coordinator]
B --> C[Planner]
C --> D[Research Team]
D --> E[Reporter]
Comparison (Pie/Bar) :
pie title Market Share
"Project A" : 45
"Project B" : 30
"Others" : 25
Assign confidence based on source quality:
| Confidence | Criteria |
|---|---|
| High (90%+) | Official docs, GitHub data, multiple corroborating sources |
| Medium (70-89%) | Single reliable source, recent articles |
| Low (50-69%) | Social media, unverified claims, outdated info |
Save report as: research_{topic}_{YYYYMMDD}.md
Weekly Installs
222
Repository
GitHub Stars
27.8K
First Seen
Feb 17, 2026
Security Audits
Gen Agent Trust HubFailSocketPassSnykWarn
Installed on
opencode218
gemini-cli217
github-copilot217
codex216
kimi-cli215
amp215
DOCX文件创建、编辑与分析完整指南 - 使用docx-js、Pandoc和Python脚本
42,600 周安装