How can I configure n8n to use search keywords and top 10 articles for content layout?

Below is one way you might implement this in n8n. While there isn’t a dedicated node that automatically uses your search keywords and the top 10 articles as titles and content, you can combine existing nodes to achieve this:

1. Fetch the Search Results
   • Use an HTTP Request node (or a custom integration if available) to query your search endpoint with your keywords.
   • Ensure the response includes article metadata (including titles, URLs, etc.).

2. Filter and Extract the Top 10 Articles
   • With a Function or Code node, you can process the search response to extract the top 10 results.
   • During this step, you can extract the article titles and any necessary content, storing them as separate objects.
   • Optionally, use the Data Filtering feature in the node’s Input or Output panels if you want a visual check of your data while developing the workflow (note that filtering here is purely visual).

3. Arrange and Use the Data
   • Once you have the top 10 articles along with your search keywords, you can use these as titles or organize the data as needed.
   • For example, if you are indexing this data in a vector store (like the Pinecone integration documented in some n8n workflows), you can incorporate the keywords into your vector data alongside the corresponding article content.

Below is an example snippet of what the Function node might look like for extracting the top 10 articles:

JAVASCRIPTCopied
// Assuming items is an array of search result objects
// and each object has a `title` and `content`

const searchKeyword = $json.searchKeyword; // your search keyword from a previous node

// Filter out the top 10 articles based on some criteria (e.g., ranking)
const topArticles = items.slice(0, 10).map(item => {
  return {
    json: {
      title: item.json.title,
      content: item.json.content,
      searchKeyword
    }
  };
});

return topArticles;

4. Further Usage
   • You can now use the output from the Function node as input for subsequent nodes.
   • For example, if you are storing data in Pinecone, the Pinecone Vector Store node can transform these articles into vectors for further querying.
   • Alternatively, you can trigger other workflows such as chat-based querying using nodes like Chat Trigger and Question and Answer Chain.

This modular approach lets you tailor your workflow to use search keywords as part of the output while arranging the content based on the top 10 articles. Although there isn’t an out-of-the-box, single-node solution for this specific functionality, combining standard nodes (HTTP Request, Function/Code, Data Filtering, etc.) makes it possible to implement tailored behavior within your n8n processes.

If you need further guidance on any specific node configurations or a particular integration, feel free to ask!