Auto-Scrape EU News and Send AI-Filtered Sustainability Daily Newsletter with n8n

Input: EU news webpage; Processing: scheduled scrape → HTML parse & split → date/type filter → GPT classify & flag → sheet log → read flagged rows → HTML email; Output: daily filtered sustainability news email

Who

Business owners, sustainability consultants, or marketing operators who need to automatically track EU sustainability developments and send daily digests to subscribers

You get

Automatically fetch EU news every morning, filter for sustainability relevance via AI, log results to Google Sheets, then read flagged rows to generate a styled HTML email to a mailing list

You need

n8n · google-sheets · gmail

Steps

1. Configure the schedule trigger

   Add a Schedule Trigger node in n8n and set it to trigger at 08:30 every morning. Adjust timezone and schedule to your needs.

2. Scrape and parse the EU news page

   Use an HTTP Request node to fetch https://commission.europa.eu/news-and-media/news_en, then use two HTML nodes: 1) Extract Articles Blocks uses CSS selector div.ecl-content-item-block__item to extract article HTML blocks (returnArray=true, returnValue=html); 2) Parse Article Blocks extracts type, date, title, link (href attribute), description, image (src attribute), and read_time from each block (see workflow.json for exact CSS selectors). Then use Split Out by Article Block to split the articles array into individual items for downstream processing.

3. Filter articles by date and type

   Use an If node to keep only articles whose date matches '5 days before today' and whose type equals News article. The date expression is {{ $now.minus(5,'day').day }} {{ $now.minus(5,'day').monthLong }} {{ $now.minus(5,'day').year }}; adjust the day offset as needed. Items passing the filter enter the Loop Over Articles node for batch processing.

4. Classify sustainability relevance with an AI Agent

   Loop Over Articles feeds single articles into the AI Agent node. The agent connects to an OpenAI Chat Model (default gpt-4o-mini) and a Structured Output Parser to constrain output format. It classifies each article based on title and description for sustainability relevance.

   Execution prompt · Source text。AI Agent node text input field (user prompt)

   Copy prompt

   =Title: {{$json.title}}
   Description: {{$json.description}}
   
   Is this article about sustainability? Return only: true or false

   Execution prompt · Source text。AI Agent node options.systemMessage field

   Copy prompt

   =You are a classification assistant. 
   
   Your role is to analyze the title and description of an article and determine if it is related to sustainability. 
   
   You must only return {"answer": true} if the article is clearly related to sustainability (e.g., environmental protection, renewable energy, sustainable development, climate action, green economy, etc.). 
   
   If it is not clearly related, return {"answer": false}.
   
   If the description is empty or missing, rely only on the title. Your response must be only one of the two JSON options: {"answer": true} or {"answer": false}. Do not provide explanations.

5. Flag and merge results

   The AI Agent output passes through Sustainability Flag (Set node, taking {{ $json.output.answer }}) to set the sustainability field, then merges with original article data via Merge Article + Flag (combineBySql mode) to form complete records.

6. Log filtered results to Google Sheets

   Use Record Results (Google Sheets node, Append mode) to add sustainability=true articles to a specified sheet. Map fields: sustainability, type, date, title, link, description, image, read_time. Configure Google Sheets API credentials and select the target file and worksheet in advance.

7. Read flagged articles and generate HTML email

   Parallel branch: the Trigger also fires Get Sustainability News (Google Sheets node, read mode, filter sustainability = true) to fetch all historically flagged articles. These feed into Generate Email HTML (Code node) to aggregate an HTML email with title, summary, image, date, and reading time. Finally, Send to your mailing list (Gmail node) delivers the email. Gmail OAuth credentials must be preconfigured; the default sendTo is [email protected] and must be replaced with actual recipients.

You are an n8n workflow setup assistant. The user wants to replicate the 'AI-filtered EU sustainability news daily newsletter' template. First confirm they have: an n8n instance, Google Sheets API credentials, Gmail OAuth credentials, and an OpenAI API key. Then assist configuration in this order: 1) import workflow.json; 2) check and update the Schedule Trigger time; 3) verify the HTTP Request target URL is accessible; 4) confirm model and credentials in the AI Agent node; 5) select the file and map fields in the Record Results node; 6) confirm read filter in the Get Sustainability News node; 7) replace the recipient address in the Gmail node and test send. Do not automatically perform any action that could send real emails or write user data.