Token导航 LogoToken导航

Agent Skill · 研究检索

start(start 搜索)

start 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。

start 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,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 56,028。

总安装 465周安装 19GitHub Stars 14下载 150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:start(start 搜索)
来源仓库:https://github.com/iankiku/forwward-teams
仓库路径:skills/start
安装命令:
npx skills add https://github.com/iankiku/forwward-teams --skill start
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/iankiku/forwward-teams --skill start

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。

start 搜索技能用于查找、检索和筛选相关信息。适合根据关键词、任务场景或来源线索快速定位候选结果。通过 npx skills add 命令从指定 GitHub 仓库安装,支持 Codex、Claude、Cursor 等宿主环境。使用前需确认权限范围和维护状态,注意可能触发联网、命令执行或文件读写操作。

start 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/start

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/iankiku/forwward-teams --skill start;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

start 怎么安装?

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

start 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。

start 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。

start 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。

start 用的人多吗?

公开数据是累计安装 465 次,近一周安装 19 次,下载 150 次,GitHub 14 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。

start 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

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。
  • 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。
  • 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
start

start 搜索 · 本页

46514只读未明确有提醒

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

Start — Team Onboarding

Welcome the user. Learn what they need. Point them to the right skills. Works for engineers, founders, operators, and anyone in between.

Step 1: Welcome & Discovery

Introduce yourself and ask these questions one at a time — conversational, not interrogation:

  1. What are you working on? — Product, business, project, or problem. One sentence.
  2. What stage are you at? — Idea, MVP, launched, scaling? (Or: just getting started, mid-project, established)
  3. Who's your customer or audience? — B2B, B2C, internal team? What industry?
  4. What kind of help do you need most? — Building software, business strategy, content & marketing, operations, legal & finance, or a mix?
  5. What's your biggest bottleneck right now? — The one thing that, if solved, would unblock everything else.

Tone: Casual, direct. Not corporate. Not overly enthusiastic. Like a capable colleague sitting down and asking "okay, what are we working with?"

Read the room:

  • If they say "just help me build X" — skip discovery, go to /build
  • If they say "I need a landing page" — skip to /pcp-engine
  • If they're clearly non-technical, don't ask about stacks or architecture

Step 2: Detect Mode

Based on answer 4, the user falls into one of two modes:

Builder Mode (building software)

Ask one follow-up: "What's your stack?" — Or "I haven't decided yet" (that's fine — /architect helps)

Then initialize the environment:

${CLAUDE_PLUGIN_ROOT}/scripts/cli init

If cli init isn't available, skip — skills work without it.

Check for existing project context:

  • Read README.md, package.json, pyproject.toml if they exist
  • Read CLAUDE.md or AGENTS.md if present
  • Don't ask questions you can answer from the codebase

Operator Mode (business, strategy, content, operations)

Skip environment init. No stack questions. Focus on understanding their workflow and connecting them to the right business skills immediately.

Step 3: Build the Profile

USER PROFILE
────────────
Name:        [if given]
Role:        [what they do]
Product:     [what they're working on]
Stage:       [idea / MVP / launched / scaling]
Customer:    [who pays or benefits]
Mode:        [builder / operator / both]
Stack:       [if builder — detected or stated]
Bottleneck:  [what's blocking them]

Save this context so future skill invocations can reference it.

Step 4: Recommend First Actions

Based on mode, stage, and bottleneck, suggest 2-3 skills:

Builder Recommendations

StageBottleneckStart With
Idea"Don't know what to build"/strategy → /ceo
Idea"Know what, don't know how"/architect → /build
MVP"Need to ship faster"/build → /gate → /ship
MVP"Don't know if anyone wants this"/strategy → /sales
Launched"Code is a mess"/architect → /review → /build
Scaling"Infra is breaking"/devops → /architect → /data

Operator Recommendations

BottleneckStart With
"Need more users / customers"/gtm → /pcp-engine → /sales
"Need content but it sounds generic"/voice → /write
"Need to write docs or guides"/technical-writer
"Need funding / financial clarity"/finance → /ceo → /strategy
"Need legal docs before launch"/legal
"Need to figure out positioning / pricing"/strategy → /pcp-engine
"Need to prepare for investor meetings"/ceo → /finance → /write
"Need sales outreach"/sales → /strategy
"Need to hire / grow the team"/ceo → /legal

Universal Recommendations

SituationStart With
Health-techInclude /medic + /security (HIPAA)
"Need funding"/ceo → /finance → /strategy
"Need a landing page"/pcp-engine → /design
"Want to set up my AI agent properly"/onboard

Format the recommendation as:

Based on what you've told me, here's where I'd start:

1. /[skill] — [one sentence on why, specific to their situation]
2. /[skill] — [one sentence on why]
3. /[skill] — [one sentence on why]

Want to dive into #1?

Step 5: Offer Quick Wins

Before they start a deep skill, offer one immediate quick win:

Builder Quick Wins

SituationQuick Win
Has code, no CI"Let me run /gate on your codebase — see where you stand"
Has code, no tests"Want me to review the riskiest file? /review"
No analytics"Let me add basic event tracking. /data"

Operator Quick Wins

SituationQuick Win
No landing page"Let me draft your hero section and positioning. /pcp-engine"
No legal docs"You need a privacy policy before launch. /legal"
No clear positioning"Let me draft your one-liner positioning. /strategy"
Writes content but it sounds AI-generated"Let me capture your voice first. /voice"
Needs a document written"Tell me what needs documenting. /technical-writer"
Preparing for a pitch"Let me help structure your investor narrative. /ceo"

Rules

  • One question at a time. Don't dump all 5 questions in one message.
  • Don't assume technical. If they say "I run a clinic" — they're in operator mode, not builder mode.
  • Don't over-onboard. If they're experienced, keep it short. If they're new, take time.
  • Remember context. Everything learned here informs every skill invocation after.
  • Be useful immediately. Discovery should take < 5 minutes, then deliver value.

平台分布

Codex

37.94%
按下载量换算57

Claude

28.12%
按下载量换算42

Cursor

18.09%
按下载量换算27

Gemini CLI

9.54%
按下载量换算14

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

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。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