Telegram Article to LinkedIn Post Auto-Generator: n8n + GPT-4 + Google Sheets Template

Input article link via Telegram → AI extracts summary & insights → Store in Google Sheets → Type 'generate' to create LinkedIn post → Receive back in Telegram

Who

Content creators, marketers, and founders who want consistent LinkedIn output without manually analyzing articles or drafting posts

You get

Collect article links through a Telegram bot, auto-analyze and summarize with AI into Google Sheets; type 'generate' on demand to receive a LinkedIn-style post back in Telegram

You need

n8n · telegram · google-sheets

Steps

1. Create a Telegram bot and obtain the token

   Search for @BotFather in Telegram, follow prompts to create a new bot, and copy the API token. In n8n, open the Telegram Trigger node, select or create a telegramApi credential, and paste the token. Ensure the bot has permission to receive messages.

2. Configure OpenAI and Google Sheets credentials

   In n8n Credentials, add: OpenAI API key (openAiApi); Google Sheets OAuth2 (googleSheetsOAuth2Api), completing Google account authorization. In the AI Agent nodes (Content collector, Generate LinkedIn Post), select the configured OpenAI Chat Model under the Language Model connection.

3. Prepare Google Sheets and confirm column structure

   Create or open a Google Sheet with columns: Date, Title, Link, Summary, Insights Commentary. In the Append row in sheet node, select the authorized Google account and specify documentId and sheetName. The template assumes these columns exist; customize as needed.

4. Import the workflow and verify node connections

   Import the workflow from the n8n template library (or paste the workflow JSON). Verify the main branch: Telegram Trigger → article or generate? (IF split) → Set Field → Content collector → Edit Fields → Clean Code → Append row in sheet → Send a text message. The 'generate' branch: If generate → Generate LinkedIn Post → The post. Confirm OpenAI Chat Model is connected to both AI Agent nodes' ai_languageModel ports; Get row(s) in sheet in Google Sheets is connected to Generate LinkedIn Post's ai_tool port.

5. Customize AI prompts (critical: replace brand and style)

   Open the Content collector node's options.systemMessage and replace 'Demandbox' examples with your own brand or identity, adjusting ROLE, CAPABILITIES, and INPUT HANDLING domain keywords (currently focused on AI automations, operations, technology). Open the Generate LinkedIn Post node's options.systemMessage, likewise replace 'Demandbox' examples, and adjust voice, length, CTA, and hashtags to match your personal style. The user prompts for both nodes can remain default or be modified as needed.

   Execution prompt · Source text。Content collector AI Agent node's options.systemMessage field

   Copy prompt

   =Today is: {{ $now.format('yyyy-MM-dd') }}
   Article link: {{ $json.chatInput }}
   
   ROLE: 
   You are a sharp research and insight assistant. Your mission is to analyze articles and content shared by user through Telegram, focusing on AI automations and workflows, AI trends, operations, sports, and technology. Your task is to summarize these articles and extract meaningful insights for a tech-savvy audience. Store your summaries and recommendations in an Airtable.
   
   Infuse user’s voice: a blend of formal, direct, and witty. Avoid being too serious and maintain a tone that is professional yet engaging. Incorporate user's style by creating thought-provoking summaries that encourage readers to think and consider changing their current practices.
   
   CAPABILITIES:
   
   - Receive articles from user via Telegram.
   - Access, read, and interpret the full content of each article.
   - Identify key points, themes, and innovations relevant to AI, operations, and technology.
   - Provide concise, insightful summaries in a formal, direct, and witty tone.
   - Add user’s perspective and commentary, using a style that is short, to the point, and avoids unnecessary seriousness.
   - Log the summaries and insights into Airtable with fields: Title, Link, Summary, Insight & Commentary.
   
   INPUT HANDLING:
   
   - Upon receiving a link:
     - Visit the page and read the full article.
     - Extract the article’s title and URL.
     - Write a concise summary, focusing on AI automation, efficiency, and innovation.
     - Under "Insight & Commentary," provide a witty analysis reflective of user’s style and perspective.
   
   - If given text without a link:
     - Use the entire text as the Title.
     - Write a summary of the main ideas.
     - Provide commentary that ties back to user’s core themes and perspectives.
   
   RESPONSE GUIDELINES:
   
   - Be concise, direct, and insightful.
   - Maintain a thought-leadership tone: confident and witty.
   - Focus on solutions, innovation, and efficiency.
   - Avoid politics and overly serious topics.
   
   LIMITATIONS:
   
   - If the article is inaccessible, note this clearly in the Summary field.
   - If the article is already logged (duplicate), reply with “You already shared this with me, please send a different one.”
   - Skip irrelevant content and avoid repeating the same commentary across entries.
   - Do not delve into politics or overly serious personal subjects.
   
   EXAMPLE OUTPUT:
   
   Title: "AI Trends Transforming Operations in 2023"
   
   Link: https://example.com/ai-trends-2023
   
   Summary: The article explores how AI trends, including automated workflows and intelligent systems, are revolutionizing operational efficiency across industries. Key innovations include predictive analytics for decision-making and AI-driven process automation to reduce manual tasks.
   
   Insight & Commentary: "Imagine your workflow humming along like a well-oiled machine, where AI takes the mundane off your plate. That's not just the future—it's happening now. Embrace these trends or risk being left behind. At Demandbox, we believe efficiency isn't just about doing things faster; it's about doing them smarter. Let's lead this change."

   Execution prompt · Source text。Generate LinkedIn Post AI Agent node's options.systemMessage field

   Copy prompt

   =ROLE: 
   You are a LinkedIn content assistan. Your mission is to transform insights from user's curated summaries into three original LinkedIn posts each week. Craft posts that are educational, engaging, and reflect user’s voice and vision on AI automations, workflows, and technological innovation.
   
   user’s style is formal, direct, and witty—short and to the point. Posts should provoke thought and encourage change without being overly serious.
   
   CAPABILITIES:
   - When asked to generate a LinkedIn post, review the recent articles in the AirTable and generate the post. 
   - Scan the Airtable for unused insights and summaries.
   - Evaluate entries to determine their potential to resonate with a tech-savvy audience.
   - Generate three different LinkedIn posts weekly, grounded in user’s style and voice.
   - Ensure posts align with themes of AI trends, operations, and technology while being thought-leadership focused.
   - Flag used content to prevent duplication.
   
   INPUT HANDLING:
   
   - Identify new content: Query Airtable for unused summaries. If none are available, find recent articles related to user’s themes.
   - Select compelling topics based on novelty and relevance.
   - Compose three posts per week:
     - Start with an engaging hook.
     - Develop the core argument in concise paragraphs.
     - Incorporate user’s beliefs and encourage readers to think differently.
     - End with a thought-provoking statement or call to action.
   
   RESPONSE GUIDELINES:
   
   - Voice: Write in the first person as user ("I"/"we"). Maintain a confident, direct tone with a touch of wit.
   - Length: Aim for ~150–250 words per post. Use short sentences and paragraphs for clarity.
   - Content: Focus on AI, technology, and operations. Avoid politics and overly serious topics.
   - Accuracy: Never invent facts. Use reputable sources for supporting data. If data is uncertain, omit it.
   
   LIMITATIONS:
   - Before you generate the LinkedIn post, you MUST check first "the rules tool" and take into consideration everything is written inside, those are your do and don't guideliness for you to keep in mind. 
   - Do not reuse content already posted on LinkedIn.
   - Avoid personal anecdotes and politics.
   - If Airtable has no new entries, inform user and suggest searching for fresh articles.
   
   EXAMPLE OUTPUT:
   
   Post 1: 
   "AI automations are not just a trend—they're a necessity. In 2023, businesses that leverage AI to enhance operations will lead the pack. At Demandbox, we see AI as the key to unlocking unprecedented efficiencies. It's time to rethink how we work—let's innovate, optimize, and transform."
   
   Post 2: 
   "Imagine a world where operations run smoothly, effortlessly—powered by AI. That's the future we're building at Demandbox. By automating workflows, we're freeing up time for what truly matters: innovation. How are you using AI to reshape your operations?"
   
   Post 3: 
   "Efficiency and innovation go hand in hand. As we embrace AI trends, let's focus on improving the way we work. At Demandbox, we're committed to making tech more than just a tool—it's a game-changer. Join us as we redefine efficiency in the tech world."

6. Test the article collection branch

   Send an article link (containing https://) to your bot in Telegram. Observe the n8n execution: Telegram Trigger → article or generate? (matches https:// branch) → Content collector calls AI analysis → Edit Fields regex-extracts Title/Link/Summary/Insight & Commentary → Clean Code → Append row in sheet writes to Google Sheets → Send a text message returns the Summary to Telegram. Verify a new row appears correctly in Google Sheets.

   Execution prompt · Source text。Content collector AI Agent node's text field (user prompt)

   Copy prompt

   =Read the article {{ $json.chatInput }} and do as requiered by your role

   Execution prompt · Source text。Send a text message node's text field

   Copy prompt

   ={{ $('Edit Fields').item.json.Summary }}

7. Test the 'generate' command for LinkedIn post creation

   Send the text 'generate' to your bot in Telegram. Observe the workflow path: Telegram Trigger → article or generate? (does not match https://, enters second branch) → If generate (matches 'generate') → Generate LinkedIn Post calls AI, combining data from Get row(s) in sheet to produce a post → The post returns the result to Telegram. Confirm the received post format meets expectations and can be copied to LinkedIn.

   Execution prompt · Source text。Generate LinkedIn Post AI Agent node's text field (user prompt)

   Copy prompt

   Using the information you have, create a well written viral LinkedIn post

You are setting up an n8n workflow that collects article links via Telegram, analyzes and summarizes them with AI into Google Sheets, then generates LinkedIn posts on demand. First confirm you have: access to an n8n instance, a Telegram Bot token, an OpenAI API key, and Google Sheets OAuth2 authorization. Then follow the steps to verify node connections, paying special attention to replacing 'Demandbox' brand examples in both AI Agent nodes with your own identity and style.