Project Access Link

https://editor.momen.app/tool/z7Bx4APAo7X/WEB?code=b9P5pHc4xubKD 

Tutorial Video

https://www.youtube.com/watch?v=a5XvuI-eSL8 

Introduction

• Goal: Automatically analyze an article’s title and content to extract 3–5 relevant keywords and save them as associated tags in the database.
• Applicable Scenario: Content management systems (CMS), blog platforms, or automated SEO tagging tools.
• Core Logic: UI Input -> Actionflow (AI Agent Analysis -> Insert Article -> Loop Insert Tags) -> Database Storage.

Steps

Data Storage

First, define the relational structure to store articles and their corresponding tags.

• Data Model:

  Table Name	Field Name	Type	Note
  article	title	Text	The headline of the article.
  	content	Text	The main body of the article.
  tag	name	Text	The keyword extracted by AI.

• Relationship: Create a One-to-Many (1:N) relationship from article to tag. One article can have multiple tags, and each tag belongs to one specific article.

Logic & State Configuration

AI Agent Configuration

Switch to the AI tab to create an agent that handles the natural language processing.

1. Inputs: Add two Text variables: title and content.
2. Prompt template:
   • Role: You are a professional Content Analysis Assistant.
   • Goals: Extract 3 to 5 core keywords/tags based on the input content. Ignore irrelevant information and ensure the tags accurately summarize the core main idea.
   • Context Reference: Reference the inputs using {{Input/title}} and {{Input/content}}.
3. Outputs: Set the type to Structured.
   • Define an object body containing an ARRAY(STRING) named tag. This ensures the AI returns a list of strings that the system can iterate through.

Actionflow Construction

Navigate to the Actionflow tab to build the backend logic.

• Name: AI Auto Tagging
• Execution Mode: Set to Async.
• Nodes:
  1. Input Node: Define title and content as input parameters.
  2. AI Node: Select the agent_smart_tagging. Bind its inputs to the Actionflow’s input parameters.
  3. Insert Data (save article): Insert a record into the article table using the inputs.
  4. Loop (loop save tags):
     • Datasource: Select the tag array returned by the AI node (AI node.body.tag).
  5. Insert Data (save tag): Inside the loop, insert a record into the tag table.
     • name: Bind to item (the current tag in the loop).
     • article_id: Bind to the id generated by the previous “save article” node.

UI Construction & Interaction

1. Component Tree:
   • Text Input (title)
   • Text Input (content)
   • Button (Get Started)
2. Interaction:
   • Select the Button.
   • Go to the Action panel -> On click -> Actionflow.
   • Choose AI Auto Tagging.
   • Data Binding:
     • title -> Text input title.value
     • content -> Text input content.value

Verification

1. Enter Preview Mode.
2. Paste a sample article title and body into the inputs.
3. Click the Get Started button.
4. Check the Data (Database) tab:
   • Verify a new record exists in the article table.
   • Open the tag table to see the keywords extracted by the AI, all correctly linked to the article’s ID.