自动化SEO审计与Slack工作流

Thumbnail for Video
AI流程的工作原理 - 自动化SEO审计与Slack工作流

流程

AI流程的工作原理

此流程中使用的提示

以下是此流程中用于实现其功能的所有提示的完整列表。提示是给予AI模型的指令,用于生成响应或执行操作。它们指导AI理解用户意图并生成相关输出。

提示词

用于自动化SEO审计和结构化SEO报告(含建议)的提示词。

                Analyze the provided page content and perform an automated SEO audit according to the checklist below, then output a structured SEO report with actionable recommendations.
Checklist & Requirements:
    On-Page Element Verification (from provided content):
        H1 tag: Verify presence, uniqueness, and inclusion of the main keyword.
        Meta title: Check length (50–60 characters), presence of main keyword, and clarity.
        Meta description: Check length (150–160 characters), presence of main keyword, and persuasiveness.
        URL length: Extract from input; ensure <75 characters, descriptive, keyword-rich.
        Image optimization: Check that each image has:
            A descriptive file name (no generic “image123.jpg”)
            Relevant alt text including the main keyword or related terms.
    Word Count:
        Count total visible text words (ignore HTML, menus, footer). Must be >800 words.
    Internal Linking:
        Detect links pointing to the domain **https://www.flowhunt.io**. Must have at least 2 links to relevant internal articles with descriptive anchor text.
    Keyword Occurrence Check:
        Verify main keyword is present in:
            H1 tag
            First 100 words of the body/excerpt
            URL slug
            Meta title

=== PAGE HTML ===
{context}
===

Output the result as a review and structure the summary of review in a markdown table with appropriate icons to draw attention if its resolved, needs review, or critical fix. make sure to include emojis for each task indicating whether its good, needs to be improved, or critical to be fixed

            

提示词

用于Google Search Console审核及博客优化建议的提示词。

                Given is the URL and content of the page. First search the URL in Google Search Console to see if this URL exists in Google Search Console. If it does exist, analyze the CTR, Position etc.. and based on the content of the webpage suggest what should be changed in the Blog and make a list of what should be changed.

=== PAGE CONTENT AND URL ===
{context}
===

Output the result as a review and structure the summary of review in a markdown table with appropriate icons to draw attention if its resolved, needs review, or critical fix. make sure to include emojis for each task indicating whether its good, needs to be improved, or critical to be fixed

            

提示词

用于同质化分析及建议的提示词。

                Given is the URL and content of the page. Your objective is to analyze the page from Cannibalization perspective. Search the queries that are ranking for the same primary keyword and see if there are any other URLs which are ranking for the same URL. If there is, suggest how to move forward and improve.

=== PAGE CONTENT AND URL ===
{context}
===

Output the result as a review and structure the summary of review in a markdown table with appropriate icons to draw attention if its resolved, needs review, or critical fix. make sure to include emojis for each task indicating whether its good, needs to be improved, or critical to be fixed

            

提示词

用于向Slack发送带有审核摘要和链接的通知消息的提示词。

                Given is link to good docs of a review for an article. the review is given as REVIEW. write a message to the SEO group in slack notifying everyone the review is done giving also an overall nice message, giving an overall feeling of the review and make sure to include the REVIEW LINK in your message so they can check it out. at the end also ask the team to answe in the same thread whether it should review again or continue with next steps to index the page.

REVIEW LINK: {context}

=== REVIEW CONTENT===
{input}
===

JUST OUTPUT THE MESSAGE THAT IS SENT TO SLACK IN MARKDOWN

            

AI Agent

拥有20年经验的SEO AI代理,执行详细的SEO审计并使用相关工具。

                Input:

Full HTML/content of a given URL. This is the source for all checks. Use the provided keyword tools only when additional keyword or SERP data is required.

Objective:

Analyze the provided page content and perform an automated SEO audit according to the checklist below, then output a structured SEO report with actionable recommendations.

Checklist & Requirements:

   On-Page Element Verification (from provided content):

       H1 tag: Verify presence, uniqueness, and inclusion of the main keyword.

       Meta title: Check length (50–60 characters), presence of main keyword, and clarity.

       Meta description: Check length (150–160 characters), presence of main keyword, and persuasiveness.

       URL length: Extract from input; ensure <75 characters, descriptive, keyword-rich.

       Image optimization: Check that each image has:

           A descriptive file name (no generic “image123.jpg”)

           Relevant alt text including the main keyword or related terms.

   Word Count:

       Count total visible text words (ignore HTML, menus, footer). Must be >800 words.

   Internal Linking:

       Detect links pointing to the same domain. Must have at least 2 links to relevant internal articles with descriptive anchor text.

   Keyword Occurrence Check:

       Verify main keyword is present in:

           H1 tag

           First 100 words of the body/excerpt

           URL slug

           Meta title

   SERP Position & CTR Review:

       If available from Google Search Console data (or supplied separately), compare ranking position for the main keyword with its CTR.

       If CTR is lower than expected for that position, suggest meta title/description tweaks.

   Cannibalization Detection:

       Use Get Keywords for Site to find keywords targeted by the domain.

       Identify if more than one URL is ranking for the same main keyword. If yes, list competing URLs and recommend consolidation/differentiation.

Available Tools:

   Get Keywords for Site – retrieve domain keywords for cannibalization checks.

   Get Keywords for Keywords – find related keywords for semantic optimization.

   Get Keywords Search Volume – verify keyword volumes and trends.

**MAKE SURE THE REPORT IS SUPER DETAILED WITH ALL DATA PRESENT AND SUGGESTIONS WITH HIGH DETAIL**

            

AI Agent

负责协调审核、使用工具获取审核内容,并根据管理员审批采取相应行动的代理。

                use the review_content_tool to fetch the review content. send as flow_variable "seo_review"

Based on the response and instructions of the user, here are the steps that you should take and respond back to the admin:

1. if the admin approves the review and approves to continue indexing the webpage, move on to use the tools you have in disposal to index the webpage in Google Search Console.

* Automatically send URL to Google Search Console Index API
* Assign priority (e.g., articles with high CPC or seasonal keywords = higher priority)

            

此流程中使用的组件

以下是此流程中用于实现其功能的所有组件的完整列表。组件是每个AI流程的构建块。它们允许您通过连接各种功能来创建复杂的交互并自动化任务。每个组件都有特定的用途,例如处理用户输入、处理数据或与外部服务集成。

聊天输入

FlowHunt 中的聊天输入组件通过在 Playground 捕获消息来启动用户交互。它作为流程的起点,使工作流能够处理文本和基于文件的输入。

FlowHunt 中的提示(Prompt)组件

了解 FlowHunt 的 Prompt 组件如何让你定义 AI 机器人的角色和行为,确保回复相关且个性化。自定义提示和模板,打造高效、具备上下文感知的聊天机器人流程。

生成器

探索 FlowHunt 的生成器组件——利用您选择的 LLM 模型进行强大的 AI 驱动文本生成。通过结合提示词、可选的系统指令,甚至图片作为输入,轻松创建动态聊天机器人回复,使其成为构建智能对话流程的核心工具。

LLM Gemini

FlowHunt 支持数十种 AI 模型,包括 Google Gemini。了解如何在 AI 工具和聊天机器人中使用 Gemini,切换模型,并控制诸如 Token 和温度等高级设置。

创建数据

创建数据组件使您能够动态生成具有可自定义字段数量的结构化数据记录。非常适合需要即时创建新数据对象的工作流,支持灵活的字段配置,并可无缝集成到其他自动化步骤中。

解析数据

解析数据组件使用可自定义的模板将结构化数据转换为纯文本。它能够灵活地格式化和转换数据输入,以用于工作流的后续环节,帮助标准化或准备信息供下游组件使用。

设置流程变量

通过设置流程变量组件,轻松管理您的工作流程,随时设置或更新变量。该工具允许您为流程中的命名变量分配值,实现动态数据处理和灵活的流程逻辑。

获取流程变量

获取流程变量组件可让您访问工作流中的变量和参数,轻松检索流程中先前存储的数据。它确保后续步骤能够获得关键信息,支持动态和具备上下文感知的自动化流程。

AI代理

FlowHunt中的AI代理组件为您的工作流赋予自主决策和工具使用能力。它利用大型语言模型并连接各种工具以解决任务、实现目标并提供智能响应。非常适合构建高级自动化和互动式AI解决方案。

MCP 客户端

使用 MCP 客户端组件轻松将多种工具集成到您的 AI 智能体中。该组件专为无缝连接设计,能够作为您的 AI 和各类外部工具之间的桥梁,提升自动化与能力,实现高级工作流。

Slack消息已接收

该组件代表与FlowHunt的Slack对话的开始,让你可以控制Flo的Slack回复被触发的时间、地点和方式。

Slack 发送消息

该组件代表 FlowHunt 向您返回的 Slack 消息。它让您可以控制 FlowHunt 消息的发送地点、方式以及通知对象。

创建 Google 文档

在自动化工作流程中轻松生成新的 Google 文档。创建 Google 文档组件允许您指定文档内容和文件名,实现文档创建与 AI 流程的无缝集成。非常适合自动化文档生成、报告和协作任务。

聊天输出

了解 FlowHunt 中的 Chat Output(聊天输出)组件——通过灵活的多部分输出完善聊天机器人回复。是实现流程顺畅结束和打造高级互动式 AI 聊天机器人的关键工具。

流程描述

目的和优势

让我们为您构建专属的AI团队

我们帮助像您这样的公司开发智能聊天机器人、MCP服务器、AI工具或其他类型的AI自动化,以在您的组织中替代人工处理重复性任务。

了解更多

SEO内容提纲生成器
SEO内容提纲生成器

SEO内容提纲生成器

通过分析给定关键词在Google上排名靠前的搜索结果,生成符合SEO要求的内容提纲。本工作流结合AI与网络搜索工具,帮助您生成结构化的博客大纲、元描述和标题,助力创作既符合搜索引擎优化又贴合受众需求的内容。...

1 分钟阅读
基于AI的SEO指标关键词研究
基于AI的SEO指标关键词研究

基于AI的SEO指标关键词研究

该工作流程通过为您的目标词生成详细的相关关键词表格(包含搜索量、CPC、竞价区间和难度等关键SEO指标),实现关键词研究自动化。非常适合希望高效拓展博客影响力并提升SEO策略的营销人员和内容创作者。...

1 分钟阅读
SEO内容差距分析器
SEO内容差距分析器

SEO内容差距分析器

此AI驱动的工作流会分析您网页的内容结构,与排名靠前的竞争对手页面进行比较,并为提升页面SEO表现提供量身定制的改进建议。通过利用谷歌搜索、先进的语言模型和智能内容提取,它帮助您弥补内容差距,优化网站以获得更高的搜索引擎排名。...

1 分钟阅读