Token导航 LogoToken导航

Agent Skill · 研究检索

ghl-ai-agents(GHLAI 特工)

ghl-ai-agents 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。

ghl-ai-agents 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 6,746。

总安装 21,077周安装 747GitHub Stars 5下载 9,505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:ghl-ai-agents(GHLAI 特工)
来源仓库:https://github.com/justin322322/ghl-skills
仓库路径:skills/ghl-ai-agents
安装命令:
npx skills add https://github.com/justin322322/ghl-skills --skill 'GHL AI Agents'
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/justin322322/ghl-skills --skill 'GHL AI Agents'

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。安装命令:npx skills add https://github.com/justin322322/ghl-skills --skill 'GHL AI Agents 建议确认权限范围、维护状态及是否会触发联网、命令执行或文件读写。

可结合原始 README 继续核验具体用法。

来源仓库内路径:skills/ghl-ai-agents

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/justin322322/ghl-skills --skill 'GHL AI Agents…;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

ghl-ai-agents 怎么安装?

在终端执行 npx skills add https://github.com/justin322322/ghl-skills --skill 'GHL AI Agents'。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

ghl-ai-agents 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。

ghl-ai-agents 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。

ghl-ai-agents 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。

ghl-ai-agents 用的人多吗?

公开数据是累计安装 21,077 次,近一周安装 747 次,下载 9,505 次,GitHub 5 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。

ghl-ai-agents 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
ghl-ai-agents

GHLAI 特工 · 本页

21,0775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

GHL AI Agents

Overview

GoHighLevel offers AI-powered automation through Voice AI (phone calls), Conversation AI (chat/SMS/social), Workflow AI (automation building), and Review AI (review management). Together, these function as a 24/7 digital employee.


AI Employee Types

TypeChannelCapabilities
Voice AIPhone calls (inbound)Answer calls, qualify leads, book appointments, collect info
Conversation AISMS, Facebook, Instagram, Chat Widget, Live ChatRespond to messages, answer FAQs, book appointments
Workflow AIWorkflow builderGenerate workflow steps using natural language prompts
Review AIReview platformsAutomate review requests and responses
Content AIEmail, SMS, FunnelsGenerate marketing copy, subject lines, content

Voice AI Setup

Prerequisites

  • Agency Pro plan or higher
  • LC Phone number (not Twilio) with A2P verification (US)
  • AI Employee feature enabled at the agency level

Step-by-Step Configuration

  1. Navigate to Settings → AI Agents
  2. Create a new agent:

- Agent Name: Descriptive name (e.g., "Front Desk AI") - Business Name: Your business name for context - Voice: Select from available voices; preview before choosing - Language/Accent: Match your target audience - Greeting: Set the initial message (e.g., "Hello, you've reached [Business]. How can I help?")

  1. Assign Phone Number:

- Go to Phone & Availability tab - Assign an LC Phone number - Configure call routing: - AI answers all calls directly, OR - AI as backup (waits for human first, picks up if no answer)

  1. Configure Agent Goals:

- Basic Mode: Simple objective selection - Advanced Mode: Detailed prompt with specific instructions

  1. Connect Calendar:

- Sync calendars for real-time appointment booking - Select which calendar the AI should use for bookings

Voice AI Prompt Engineering

Basic Guidelines
You are a friendly, professional receptionist for [Business Name].
Your primary goals are:
1. Greet the caller warmly
2. Identify their reason for calling
3. Book an appointment if appropriate
4. Collect their name, email, and phone number
5. If you can't help, offer to transfer to a team member

Important rules:
- Never make up information you don't know
- Always confirm details before booking
- Keep responses concise and natural
- If asked about pricing, say "[specific response]"
- For emergencies, direct them to [number/action]
Advanced Prompt Template
## Identity
You are [Name], the virtual assistant for [Business Name].
Personality: [Professional / Friendly / Casual]
Tone: [Warm and helpful / Direct and efficient]

## Goals (in priority order)
1. [Primary goal — e.g., Book a consultation]
2. [Secondary goal — e.g., Collect contact info]
3. [Tertiary goal — e.g., Answer FAQs]

## Knowledge
- Services offered: [list]
- Business hours: [hours]
- Location: [address]
- Pricing: [how to handle pricing questions]

## Handling Specific Scenarios
- If caller asks about [topic]: [specific response]
- If caller is upset: [de-escalation approach]
- If caller asks to speak to a human: [transfer/callback protocol]
- If outside business hours: [after-hours message]

## Restrictions
- DO NOT: [list things the AI should never do]
- ALWAYS: [list things the AI must do]

Conversation AI Setup

Supported Channels

  • SMS / MMS
  • Facebook Messenger
  • Instagram DMs
  • Website Chat Widget
  • Live Chat
  • Google Business Messages

Configuration Steps

  1. Navigate to Settings → Conversation AI
  2. Enable channels where the bot should operate
  3. Set mode:

- Autopilot: Fully automated responses - Suggestive: AI suggests responses for human approval

  1. Configure response settings:

- Wait time before responding - Maximum messages per conversation - Handoff rules (when to transfer to human)

Training the Knowledge Base

The Conversation AI learns from your provided content to answer questions accurately.

Training Sources:

SourceHow
Website URLPaste your site URL; AI crawls and learns the content
Google DocsLink docs with FAQs, service descriptions, policies
Manual Q&AAdd specific question-answer pairs directly
File UploadUpload PDFs, documents with business information

Best Practices:

  1. Start with FAQs — Add the 20 most common questions and ideal answers
  2. Be specific — Vague training data produces vague responses
  3. Update regularly — Add new Q&As as you discover gaps
  4. Test thoroughly — Ask the bot questions from different angles
  5. Review conversations — Check chat logs for incorrect or missed responses

Appointment Booking via Chat

  1. Enable Appointment Booking in Conversation AI settings
  2. Select the target calendar
  3. The AI will:

- Suggest available time slots - Confirm the booking details - Create the appointment in the calendar - Send confirmation to the lead


Content AI

Use built-in AI for content generation across GHL:

FeatureWhereWhat It Does
Email CopyEmail BuilderGenerate subject lines, body copy, CTAs
SMS CopySMS ActionDraft text messages
Funnel CopyPage BuilderWrite headlines, descriptions, bullet points
Social PostsSocial PlannerGenerate social media content
Workflow PromptsWorkflow BuilderCreate workflow steps from natural language

AI Best Practices

General

  1. Set clear boundaries — Define exactly what the AI can and cannot do
  2. Start conservative — Begin with limited scope, expand as you gain confidence
  3. Monitor regularly — Review AI conversations daily during the first week
  4. Have a human fallback — Always provide a way to reach a real person
  5. Test edge cases — Ask unusual or adversarial questions to find weak spots

Prompt Engineering Tips

  1. Be literal — The AI follows instructions exactly; be explicit
  2. Use examples — Show desired input/output pairs
  3. Define the persona — Give the AI a name, role, and personality
  4. Set guardrails — Clearly state topics to avoid or redirect
  5. Iterate — Refine prompts based on real conversation analysis

Performance Optimization

  1. Track booking rate — Measure how many conversations result in appointments
  2. Monitor CSAT — Collect satisfaction feedback after AI interactions
  3. Analyze drop-offs — Identify where conversations end without conversion
  4. A/B test prompts — Try different approaches and measure results

Key Resources

平台分布

Codex

37.35%
按下载量换算3,550

Claude

27.25%
按下载量换算2,590

Cursor

17.54%
按下载量换算1,667

Gemini CLI

8.69%
按下载量换算826

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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