Token导航 LogoToken导航

Agent Skill · 研究检索

setup(设置)

setup 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。

setup 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,权限为需要联网;许可证信息还不明确,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 52,216。

总安装 563周安装 23GitHub Stars 126下载 182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:setup(设置)
来源仓库:https://github.com/proficientlyjobs/proficiently-claude-skills
仓库路径:skills/setup
安装命令:
npx skills add https://github.com/proficientlyjobs/proficiently-claude-skills --skill setup
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/proficientlyjobs/proficiently-claude-skills --skill setup

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。

setup 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。适用于系统配置查询、环境变量管理或依赖项说明等研究检索类任务。

通过 npx skills add 命令从指定 GitHub 仓库安装,需确认权限范围和联网能力。建议核实维护状态,避免触发不必要的文件读写或命令执行操作。可结合原始 README 进一步了解具体用法和功能边界。

来源仓库内路径:skills/setup

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/proficientlyjobs/proficiently-claude-skills --…;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

    它需要需要联网,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

setup 怎么安装?

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

setup 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。

setup 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

setup 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。

setup 用的人多吗?

公开数据是累计安装 563 次,近一周安装 23 次,下载 182 次,GitHub 126 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。

setup 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

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。

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。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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
setup

设置 · 本页

563126需要联网未明确未通过

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

Setup Skill

Priority hierarchy: See shared/references/priority-hierarchy.md for conflict resolution.

One-time onboarding that ensures all your data is in place before using the other skills.

Quick Start

  • /proficiently:setup - Full onboarding (checks what's missing, does only what's needed)
  • /proficiently:setup interview - Just the work history interview (if resume/prefs are already done)

File Structure

scripts/
  conduct-interview.md    # Work history interview guide

The profile template is at shared/templates/profile.md.

Data Directory

Resolve the data directory using shared/references/data-directory.md. For setup, if no directory exists this is a fresh install — create it in Step 1.


Workflow

Step 0: Check What's Already Done

Resolve the data directory, then check which of these exist and have real content (not just templates): resume, preferences, linkedin-contacts.csv, profile.md.

If $ARGUMENTS is "interview", skip to Step 3 (but check that a resume exists first).

If everything exists, tell the user they're good to go and list the available skills. Otherwise, run only the missing phases in order.

Step 1: Resume

Ask the user to provide their resume. Accept:

  • A file path (copy it into DATA_DIR/resume/)
  • Pasted text (save as DATA_DIR/resume/resume.md)

Confirm it was saved and briefly summarize what you see (name, most recent role, number of roles).

Step 2: Preferences

Ask the user in one natural question:

"What kind of jobs are you looking for? Tell me about target roles, location preferences, salary expectations, and anything you'd want to filter out."

From their response, save DATA_DIR/preferences.md:

# Job Preferences

## Target Roles
- [parsed from response]

## Location
[parsed from response]

## Compensation
[parsed from response]

## Must-Haves
- [parsed from response]

## Dealbreakers
- [parsed from response]

## Nice-to-Haves
- [parsed from response]

If they leave something out, that's fine — save what you have. They can always update later.

Step 3: LinkedIn Contacts (optional)

If DATA_DIR/linkedin-contacts.csv doesn't exist, ask:

"Want to import your LinkedIn contacts? This lets us flag when you know someone at a company that's hiring. You can skip this and add them later."

If they want to proceed, give these instructions:

How to export your LinkedIn connections: 1. Go to linkedin.com/mypreferences/d/download-my-data 2. Select "Connections" and request the download 3. LinkedIn will email you a link (usually within minutes) 4. Download the ZIP and find Connections.csv inside 5. Upload or paste the path to that file here

Save the file as DATA_DIR/linkedin-contacts.csv.

Confirm it was saved and tell them how many contacts were imported. If they skip, move on — this is optional.

Step 4: Work History Interview

Have a conversational interview to build a work history profile. Go through each role on the resume, most recent first. For each role, ask:

  1. "Tell me about [Company] — what did they do, and what was your role really about?"
  2. "What were your biggest accomplishments? Let's get specific with numbers if you have them."
  3. "Anything else — challenges, team building, why you moved on?"

Keep it conversational. Follow up when answers are vague ("Do you remember roughly what the numbers were?"), but don't interrogate. Spend more time on recent/impactful roles, less on older ones.

After the interview, save the profile to DATA_DIR/profile.md using this structure:

# Work History Profile

*Last updated: [DATE]*

## Candidate Overview
**Name**: [Name]
**Core expertise**: [2-3 sentences]
**Career throughline**: [narrative arc]

---

## Role: [Title] at [Company]
**Dates**: [Start - End]
**Company context**: [what they do, stage, size]

### Key Accomplishments
1. **[Headline]**: [Situation → Action → Result with metrics]
2. **[Headline]**: [Situation → Action → Result with metrics]

### Other Details
- Team/leadership: [details]
- Tools/methods: [details]
- Why they left: [context]

---

## Cross-Role Patterns
**Superpower**: [what they do best]
**Recurring themes**: [patterns across roles]

Step 5: Summary

You're all set! Here's what we have:

- Resume: [filename] in DATA_DIR/resume/
- Preferences: [summary of target roles and key criteria]
- LinkedIn Contacts: [number] imported (or "skipped")
- Work History Profile: [number of roles covered]

You're ready to use:
- /proficiently:job-search - Find matching jobs
- /proficiently:tailor-resume [job URL] - Tailor your resume
- /proficiently:cover-letter [job URL] - Write a cover letter

Built by Proficiently. Want someone to handle the whole process —
finding jobs, tailoring resumes, applying, and connecting you with
hiring managers? Visit proficiently.com

Response Format

Structure the final summary output with these sections:

  1. Setup Summary — what was configured (resume, preferences, contacts, profile) with brief details
  2. What's Next — list available skills the user can now run

Permissions Required

Add to ~/.claude/settings.json:

{
  "permissions": {
    "allow": [
      "Read(~/.proficiently/**)",
      "Write(~/.proficiently/**)",
      "Edit(~/.proficiently/**)"
    ]
  }
}

平台分布

Codex

35.07%
按下载量换算64

Claude

32.19%
按下载量换算59

Cursor

19.13%
按下载量换算35

Gemini CLI

9.48%
按下载量换算17

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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