Token导航 LogoToken导航

Agent Skill · 研究检索

onboard(新手引导设计)

onboard 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。

onboard 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;许可证为 MIT,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 73,530。

总安装 20周安装 17GitHub Stars 1,390下载 140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
MIT

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:onboard(新手引导设计)
来源仓库:https://github.com/ballred/obsidian-claude-pkm
仓库路径:skills/onboard
安装命令:
npx skills add https://github.com/ballred/obsidian-claude-pkm --skill onboard
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/ballred/obsidian-claude-pkm --skill onboard

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。

onboard 用于查找、检索和筛选相关信息。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词快速定位候选结果。通过 npx skills add 命令从指定仓库安装,需确认权限和维护状态。使用前建议核实是否会触发联网、命令执行或文件读写操作。

可结合原始 README 进一步验证具体用法和功能边界。

来源仓库内路径:skills/onboard

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/ballred/obsidian-claude-pkm --skill onboard;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

onboard 怎么安装?

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

onboard 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。

onboard 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。

onboard 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

onboard 用的人多吗?

公开数据是累计安装 20 次,近一周安装 17 次,下载 140 次,GitHub 1,390 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。

onboard 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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 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

  • 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。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 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
onboard

新手引导设计 · 本页

201,390需要联网MIT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

Onboard Skill

Interactive vault setup (first run) and context loading (subsequent runs).

Usage

/onboard                    # Full onboard (setup if first run, context load if not)
/onboard Projects/MyProject # Load specific project context

First-Run Setup

If the file FIRST_RUN exists in the vault root, this is a new vault. Run the interactive setup:

Step 1: Welcome

Greet the user and explain what will happen:

  • "I'll ask a few questions to personalize your vault (~2 minutes)"
  • "Your answers are saved locally in vault-config.json"
  • "You can change these anytime by editing that file or running /onboard again"

Step 2: Ask Questions

Use AskUserQuestion to ask these interactively:

Question 1: Your name

  • "What should I call you?"
  • Used for personalized prompts and greetings

Question 2: Preferred review day

  • "What day do you prefer for your weekly review?"
  • Options: Sunday (Recommended), Saturday, Monday, Friday
  • Used by /review auto-detection and session-init nudges

Question 3: Primary goal areas

  • "Which areas are most important to you right now? (Pick 2-4)"
  • Options: Career & Professional, Health & Wellness, Relationships, Personal Growth
  • Also offer: Financial, Creativity & Fun, Learning, Other
  • multiSelect: true
  • Used to customize goal template suggestions

Question 4: Work style

  • "How do you prefer Claude to interact?"
  • Options: Direct and concise (Recommended), Coaching and challenging, Detailed and thorough, Minimal — just do the task
  • Sets output style preference

Step 3: Save Configuration

Write vault-config.json in the vault root:

{
  "name": "User's name",
  "reviewDay": "Sunday",
  "goalAreas": ["Career & Professional", "Health & Wellness"],
  "workStyle": "Direct and concise",
  "setupDate": "2026-02-15",
  "version": "3.1"
}

Step 4: Personalize CLAUDE.md

Edit the root CLAUDE.md:

  • Replace [CUSTOMIZE: Add your personal mission statement here] with a prompt based on their goal areas
  • Update the "Current Focus" section to reference their chosen areas

Step 5: Remove First-Run Marker

rm FIRST_RUN

Step 6: Confirm Setup

Tell the user:

  • "Your vault is set up! Here's what's available:"
  • Brief cascade overview
  • "Try /daily to start your first morning routine"
  • "Try /review anytime — it auto-detects the right review type"

Then proceed to the standard context loading below.

Standard Context Loading (Subsequent Runs)

What This Skill Does

  1. Discovers Context Files

- Searches for all CLAUDE.md files - Traverses project directories - Respects depth limits

  1. Loads Hierarchical Context

- Root CLAUDE.md first (global context) - Project-specific CLAUDE.md files - Recent daily notes for current state

  1. Scans Active Projects

- Auto-discovers all Projects/*/CLAUDE.md files - Extracts project name, phase, progress, and goal linkage - Displays active project count and summary in onboard output

  1. Reads User Preferences

- Loads vault-config.json if present - Applies name, review day, work style preferences - Uses goal areas to prioritize context loading

  1. Builds Understanding

- Your personal mission/goals - Project structures and status - Workflow preferences - Custom conventions

Context Hierarchy

vault/
├── CLAUDE.md                 # [1] Global context - loaded first
├── Projects/
│   ├── Project A/
│   │   └── CLAUDE.md         # [2] Project context
│   └── Project B/
│       └── CLAUDE.md         # [3] Another project context
└── Areas/
    └── Health/
        └── CLAUDE.md         # [4] Area-specific context

CLAUDE.md File Structure

Root CLAUDE.md Should Include

# System Context for Claude

## Personal Mission
[Your life mission/purpose]

## Current Focus
[What you're working on now]

## Preferences
- Writing style: [Formal/Casual/Technical]
- Detail level: [High/Medium/Low]

## Conventions
- File naming: [Your patterns]
- Tag system: [Your tags]

Project CLAUDE.md Should Include

# Project: [Name]

## Overview
[What this project is about]

## Current Status
[Where things stand]

## Key Decisions
[Important choices made]

## Next Steps
[What needs to happen]

Smart Context Loading

Recent Activity

Automatically considers:

  • Last 7 days of daily notes
  • Current week's review
  • Recently modified projects

Project Summary

When loading full context, include a project overview:

### Active Projects (N)
| Project | Phase | Progress | Goal |
|---------|-------|----------|------|
| [[ProjectA]] | Active | 60% | [[Goal 1]] |
| [[ProjectB]] | Planning | 10% | [[Goal 3]] |

Selective Loading

For focused assistance:

/onboard Projects/WebApp      # Only specific project
/onboard Goals                # Only goals context

Use Cases

Starting a Session

/onboard
"Help me plan my day based on my goals"

Project Work

/onboard Projects/MyApp
"Help me refactor the authentication module"

Weekly Planning

/onboard Goals
"Analyze my week and suggest improvements"

Context Variables

Your CLAUDE.md files can include preferences:

## Variables for Claude
- DEFAULT_LANGUAGE: JavaScript
- TIMEZONE: America/New_York
- COMMUNICATION_STYLE: Direct and concise

Best Practices

Keep Context Updated

  • Review CLAUDE.md files monthly
  • Update after major decisions
  • Remove outdated information
  • Add new learnings

Be Specific

  • Clear project descriptions
  • Specific preferences
  • Concrete examples
  • Defined conventions

Hierarchical Information

  • Global → Area → Project → Task
  • General → Specific
  • Strategic → Tactical

Privacy & Security

Never Include in CLAUDE.md

  • Passwords or credentials
  • Personal identification numbers
  • Financial account details
  • Private API keys

Safe Context Examples

  • "I work in healthcare technology"
  • "My projects involve web development"
  • "I prefer morning work sessions"

Integration

Works with:

  • All other skills (provides context)
  • /daily - Better daily planning with context
  • /weekly - Informed weekly reviews
  • /monthly - Monthly review with full context
  • /project - Project status overview
  • Goal tracking - Understand goal cascade

平台分布

Claude Code

28.46%
按下载量换算40

trae

21.67%
按下载量换算30

Gemini CLI

18.24%
按下载量换算26

Antigravity

11.77%
按下载量换算16

windsurf

8.22%
按下载量换算12

Codex

3.76%
按下载量换算5

安全审计

Gen Agent Trust Hub

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。

继续浏览同类 Skills