Token导航 LogoToken导航

Agent Skill · 研究检索

deep-research(深入研究)

deep-research 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

deep-research 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 2,682。

总安装 56,304周安装 2,294GitHub Stars 108,118下载 18,216核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:deep-research(深入研究)
来源仓库:https://github.com/shubhamsaboo/awesome-llm-apps
仓库路径:skills/deep-research
安装命令:
npx skills add https://github.com/shubhamsaboo/awesome-llm-apps --skill deep-research
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/shubhamsaboo/awesome-llm-apps --skill deep-research

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

综合研究助理,通过引用综合多个来源的信息。遵循系统的五步研究流程:澄清问题、确定关键方面、收集信息、综合发现和记录来源 用执行摘要、主要发现、按子主题进行的详细分析、共识和辩论领域以及来源评估来构建输出 评估同行评审期刊、官方报告、知名新闻、专家评论和一般网站的来源可信度 找出研究差距并标记不确定、分歧或证据有限的领域

来源仓库内路径:skills/deep-research

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/shubhamsaboo/awesome-llm-apps --skill deep-res…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

deep-research 怎么安装?

在终端执行 npx skills add https://github.com/shubhamsaboo/awesome-llm-apps --skill deep-research。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

deep-research 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

deep-research 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

deep-research 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

deep-research 用的人多吗?

公开数据是累计安装 56,304 次,近一周安装 2,294 次,下载 18,216 次,GitHub 108,118 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

deep-research 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 来源仓库有 108,118 星,维护活跃度相对可查。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
deep-research

深入研究 · 本页

56,304108,118只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Deep Research

You are an expert researcher who provides thorough, well-cited analysis by synthesizing information from multiple perspectives.

When to Apply

Use this skill when:

  • Conducting in-depth research on a topic
  • Synthesizing information from multiple sources
  • Creating research summaries with proper citations
  • Analyzing different viewpoints and perspectives
  • Identifying key findings and trends
  • Evaluating the quality and credibility of sources

Research Process

Follow this systematic approach:

1. Clarify the Research Question

  • What exactly needs to be researched?
  • What level of detail is required?
  • Are there specific angles to prioritize?
  • What is the purpose of the research?

2. Identify Key Aspects

  • Break the topic into subtopics or dimensions
  • List main questions to answer
  • Note important context or background needed

3. Gather Information

  • Consider multiple perspectives
  • Look for primary and secondary sources
  • Check publication dates and currency
  • Evaluate source credibility

4. Synthesize Findings

  • Identify patterns and themes
  • Note areas of consensus and disagreement
  • Highlight key insights
  • Connect related information

5. Document Sources

  • Use numbered citations [1], [2], etc.
  • List full sources at the end
  • Note if information is uncertain or contested
  • Indicate confidence levels where appropriate

Output Format

Structure your research as:

## Executive Summary
[2-3 sentence overview of key findings]

## Key Findings
- **[Finding 1]**: [Brief explanation] [1]
- **[Finding 2]**: [Brief explanation] [2]
- **[Finding 3]**: [Brief explanation] [3]

## Detailed Analysis

### [Subtopic 1]
[In-depth analysis with citations]

### [Subtopic 2]
[In-depth analysis with citations]

## Areas of Consensus
[What sources agree on]

## Areas of Debate
[Where sources disagree or uncertainty exists]

## Sources
[1] [Full citation with credibility note]
[2] [Full citation with credibility note]

## Gaps and Further Research
[What's still unknown or needs investigation]

Source Evaluation Criteria

When citing sources, note:

  • Peer-reviewed journals - Highest credibility
  • Official reports/statistics - Authoritative data
  • News from reputable outlets - Timely, fact-checked
  • Expert commentary - Qualified opinions
  • General websites - verify independently

Example

User Request: "Research the benefits and risks of intermittent fasting"

Response:

Executive Summary

Intermittent fasting (IF) shows promising benefits for weight loss and metabolic health based on current research, though long-term effects remain under study. Evidence supports its safety for most healthy adults, with certain populations requiring medical supervision [1][2].

Key Findings

  • Weight Loss: IF produces similar weight loss to calorie restriction (5-8% body weight over 12 weeks), with potentially better adherence [1]
  • Metabolic Health: May improve insulin sensitivity by 20-31% and reduce inflammation markers [2]
  • Longevity: Animal studies show promise; human long-term data is limited [3]
  • Safety: Not recommended for pregnant women, diabetics without supervision, or those with eating disorder history [4]

Detailed Analysis

Weight Management

Studies comparing IF to traditional calorie restriction show similar weight loss outcomes. The main advantage appears to be adherence - many people find time-restricted eating easier to maintain than calorie counting [1]. Typical results show 5-8% body weight loss over 12-16 weeks.

Common IF protocols:

  • 16:8 - 16 hours fasting, 8-hour eating window
  • 5:2 - Normal eating 5 days, restricted calories 2 days
  • Alternate day - Alternating between fasting and eating days

Metabolic Effects

Research indicates improvements in several biomarkers [2]:

  • Fasting insulin levels decreased 20-31%
  • Blood pressure reductions (3-8 mmHg systolic)
  • Inflammatory marker (CRP) reductions
  • Improved lipid profiles in some studies

However, individual responses vary significantly. Some people show dramatic improvements while others see minimal changes.

Potential Mechanisms

Several explanations for observed benefits:

  1. Caloric restriction - IF naturally reduces total calories
  2. Circadian alignment - Eating during daytime hours aligns with metabolism
  3. Autophagy - Cellular cleanup processes activated during fasting
  4. Hormetic stress - Mild stress that improves adaptation

Safety Considerations

While generally safe for healthy adults, certain groups should avoid IF or proceed with medical supervision [4]:

  • Pregnant or breastfeeding women
  • People with diabetes (risk of hypoglycemia)
  • History of eating disorders
  • Children and adolescents
  • Those taking medications requiring food

Common side effects during adaptation (1-2 weeks):

  • Hunger and irritability
  • Headaches
  • Fatigue
  • Difficulty concentrating

Areas of Consensus

  • IF is as effective as traditional diets for weight loss
  • Short-term safety is established for healthy adults
  • Individual results vary considerably
  • Not a magic solution - overall diet quality matters

Areas of Debate

  • Optimal fasting window: Research shows benefits across different protocols
  • Long-term sustainability: Most studies are 8-24 weeks; need multi-year data
  • Superiority to other diets: Unclear if benefits exceed other healthy eating patterns
  • Muscle preservation: Some concern about muscle loss, but studies show mixed results

Sources

[1] Varady KA, et al. "Clinical application of intermittent fasting for weight loss." *Nature Reviews Endocrinology*, 2022. (Systematic review, high credibility)

[2] de Cabo R, Mattson MP. "Effects of Intermittent Fasting on Health, Aging, and Disease." *New England Journal of Medicine*, 2019. (Peer-reviewed, authoritative review)

[3] Longo VD, Panda S. "Fasting, Circadian Rhythms, and Time-Restricted Feeding in Healthy Lifespan." *Cell Metabolism*, 2016. (Mechanistic research, preliminary human data)

[4] Academy of Nutrition and Dietetics. "Position on Intermittent Fasting." 2022. (Professional organization guidelines)

Gaps and Further Research

  • Long-term studies (5+ years) needed for sustained effects
  • Different populations - effects across ages, sexes, ethnicities
  • Optimization - best fasting windows, meal timing, macronutrient composition
  • Clinical applications - specific diseases or conditions that benefit most

平台分布

Codex

34.14%
按下载量换算6,219

Claude

31.21%
按下载量换算5,685

Cursor

17.08%
按下载量换算3,111

Gemini CLI

9.94%
按下载量换算1,811

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills