Token导航 LogoToken导航

Agent Skill · 研究检索

research(研究)

用于辅助数据整理、表格处理、CSV/Excel 分析、指标计算和图表准备。它适合让 Agent 清洗字段、汇总数据、发现异常、生成统计口径或把分析结果转成可读说明。使用时需要确认数据来源、字段含义和时间范围,避免把样本数据当全量事实;涉及敏感数据、导出文件或批量写回时,应先确认权限和脱敏边界。

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,权限为操作浏览器;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 70,831。

总安装 199周安装 8GitHub Stars 1下载 65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:research(研究)
来源仓库:https://github.com/olamedia/analytics-skills
仓库路径:skills/research
安装命令:
npx skills add https://github.com/olamedia/analytics-skills --skill 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/olamedia/analytics-skills --skill 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。

用于辅助数据整理、表格处理、CSV/Excel 分析和指标计算。适合清洗字段、汇总数据、发现异常或生成统计口径与可读说明。需确认数据来源、字段含义和时间范围,避免将样本当全量事实。涉及敏感数据或批量操作时应先确认权限和脱敏边界。

安装命令:npx skills add https://github.com/olamedia/analytics-skills --skill research

来源仓库内路径:skills/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/olamedia/analytics-skills --skill research;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

    它需要操作浏览器,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

research 怎么安装?

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

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。

research 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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。

research 用的人多吗?

公开数据是累计安装 199 次,近一周安装 8 次,下载 65 次,GitHub 1 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。

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。

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。

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
research

研究 · 本页

1991操作浏览器未明确通过

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

Research

Scan a codebase and produce a structured context map that all subsequent analytics skills consume.

When to Use

  • Starting analysis on a project for the first time
  • Before running goal-definition, brainstorming, prd, architecture, or breakdown
  • When switching to a different project or codebase
  • When the codebase has changed significantly since the last context map

When NOT to use: The project already has a recent, accurate context-map.md in the artifact folder.

Input

  • A codebase or project path (provided by user or detected from workspace)
  • Check references/context-sources.md for known project documentation files (docs/TechStack.md, docs/ProjectStructure.md). If they exist and are fresh, use them to pre-populate tech stack and structure findings — then focus scanning on patterns, conventions, rules, and health notes.

Output

  • context-map.md saved to the artifact folder (see references/formats.md for template)

The Process

Step 1: Determine Artifact Folder

If this is the first skill called in the pipeline, ask the user where to save artifacts:

"Where should I save the analysis artifacts for this feature? Provide a folder path (e.g. docs/analytics/user-auth/), or I'll use docs/analytics/ as default."

If the user specifies a folder in chat, use that. Create the folder if it doesn't exist.

Step 2: Scan Project Structure

Use Glob to map the directory tree. Identify:

  • Source code directories and their organization
  • Test directories and testing approach
  • Configuration files (package.json, tsconfig, vite.config, etc.)
  • Documentation directories
  • Build and deployment configuration

Step 3: Identify Tech Stack

Read key configuration files to determine:

  • Languages and versions
  • Frameworks and libraries
  • Database technology
  • Build tools and dev server
  • Testing frameworks
  • CI/CD setup

Step 4: Discover Patterns and Conventions

Use Grep and Read to examine representative files:

  • Read 2-3 component/module files to identify structural patterns
  • Read 1-2 test files to understand testing approach
  • Read API/route files to understand backend patterns
  • Look for state management patterns
  • Check naming conventions (files, functions, variables, CSS classes)
  • Check error handling patterns

Step 5: Find Rules Files

Search for existing rules and convention files:

  • CLAUDE.md, .cursorrules, .cursor/rules/*.md
  • AGENTS.md, .windsurfrules
  • .editorconfig, .eslintrc, .prettierrc
  • CONTRIBUTING.md, style guides

Summarize key contents from any found.

Step 6: Note Constraints and Health

Identify:

  • Technical constraints (browser support, API compatibility, etc.)
  • Large or complex files that may need attention
  • Missing test coverage areas
  • Unclear boundaries between modules
  • Any existing features or patterns relevant to upcoming work

Step 7: Write Context Map

Write context-map.md to the artifact folder using the template from references/formats.md. Include all findings from Steps 2-6.

Announce the saved path:

"Context map saved to [path]/context-map.md."

Common Rationalizations

RationalizationReality
"I already know this codebase"Context maps surface patterns you take for granted. Write it down for the skills that follow.
"This is a small project, I don't need a map"Small projects still have conventions and constraints. A short context map is fine.
"I'll just read files as I go"Downstream skills need structured context upfront. A context map prevents repeated scanning.
"The README has everything"READMEs describe intent. Context maps describe reality — what's actually in the code.

Red Flags

  • Skipping the scan and writing a context map from memory
  • Not reading actual source files (only config files)
  • Missing the tech stack or key patterns
  • Not checking for existing rules files
  • Context map that doesn't mention any specific file paths

Verification

Before handing off, confirm:

  • Artifact folder exists and path is known
  • Tech stack identified with versions
  • Project structure documented with key directories
  • At least 3 patterns/conventions identified from actual code
  • Rules files searched for and documented (even if none found)
  • Constraints and health notes included
  • context-map.md saved to artifact folder

Next

"Research complete. Next recommended skill: goal-definition — define what you want to achieve using this context."

平台分布

Codex

32.93%
按下载量换算21

Claude

30.92%
按下载量换算20

Cursor

21.11%
按下载量换算14

Gemini CLI

9.13%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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