Automate Hacker News Summaries to AI-Driven Video Creation with n8n

Opening Problem Statement

Meet Alex, a content creator and AI automation enthusiast who spends hours daily sifting through Hacker News articles to find relevant stories about AI and automation. Alex wants to share concise, compelling video summaries but struggles with the time-consuming process of manual research, summary writing, image generation, video editing, and social media posting. This leads to missed opportunities for engagement and wasted creative energy.

Specifically, Alex often wastes 4-5 hours per day manually creating content from scattered articles. Errors in summarizing and delays in posting impact the quality and timing, resulting in lost audience growth and lower social media reach. Clearly, Alex needs a streamlined, automated approach tailored exactly to the task of turning Hacker News content into engaging videos.

What This Automation Does ⚙️

This powerful n8n workflow automates the entire process from fetching Hacker News articles to generating AI-enhanced short videos ready for social sharing. Here’s what happens step-by-step when it runs:

• Fetches latest articles from Hacker News automatically.
• Filters articles for relevance to AI and automation topics using OpenAI GPT-4 analysis.
• Summarizes relevant articles into concise newsletter-style blurbs with creative prompts for image generation.
• Generates images using Leonardo.ai based on the AI-crafted image prompts.
• Creates dynamic videos from images with RunwayML and Creatomate APIs, adding voiceovers generated via OpenAI text-to-speech.
• Uploads and posts final assets to cloud storage (Minio, Dropbox, Google Drive, OneDrive) and social media platforms including YouTube, Twitter (X), Instagram, and LinkedIn.

The automation can handle up to 50 articles per run, saving Alex countless hours weekly, eliminating manual errors, and ensuring consistent, high-quality, multi-platform content production. This workflow is a game changer for any content creator looking to leverage AI-driven automation in video content.

Prerequisites ⚙️

• 🔑 n8n Account: An active n8n instance (cloud or self-hosted). See self-hosting options for cost-effective control.
• 🔑 OpenAI API Key: For GPT-4-based article analysis and text-to-speech generation.
• 🔑 Leonardo.ai Account: To generate high-quality images from AI prompts.
• 🔑 RunwayML API Access: For video creation from generated images.
• 🔑 Creatomate API Key: For stitching videos with voiceover and subtitles.
• 🔑 Cloud Storage Providers: Minio, Dropbox, Google Drive, Microsoft OneDrive for asset uploads.
• 🔑 Social Media Accounts: YouTube, Twitter/X, Instagram (via HTTP), LinkedIn for video posting.

Step-by-Step Guide to Build This Workflow ✏️

Step 1: Start with Manual Trigger

Go to your n8n editor. Click + Create → add a Manual Trigger node. This node allows you to run the workflow on demand for testing. Position it at the start.

Expected outcome: Clicking “Execute Workflow” triggers the process.

Common mistake: Forgetting to connect the next node.

Step 2: Fetch Hacker News Articles

Add the built-in Hacker News node next. Set resource to all (default). Connect from Manual Trigger to Hacker News node.

This gets the latest articles as JSON objects.

Step 3: Limit Batch Size

Add a Limit node to restrict the max articles to 50. Connect Hacker News output to Limit input.

Why: Keeps workflow manageable for processing and API rate limits.

Step 4: Loop Over Each Article

Connect the Limit node to the Split Into Batches node (named Loop Over Items). This splits the batch into individual items for sequential processing.

Step 5: Analyze Article Relevance and Summary with OpenAI

Add the OpenAI Chat Model3 node from LangChain. It receives each article URL through an HTTP Request node that fetches the raw article content.

Configuration:

• Use HTTP Request node (Tool HTTP Request) to load article content dynamically using {{$json.url}}
• Connect the HTTP Request node’s output to OpenAI Chat node.
• OpenAI node has a system prompt to check if article is about automation or AI.
• Request a 250-word summary, image URL list, and relevance yes/no.

Step 6: Parse AI Output Structurally

Use the Structured Output Parser node to convert raw AI JSON into structured fields: summary, related (yes/no), and image URLs.

Step 7: Filter by Topic

Add the If node (named If Topic). Configure it to continue only if the “related” field equals “yes” (meaning related to AI or automation).

If not related, the article is skipped, and the workflow loops to next item.

Step 8: Image Analysis

Pass image URLs to Image Analysis (OpenAI) node to analyze and confirm image content aligns with article theme.

Step 9: Prepare Article for Content Generation

Use the Article Prep node (OpenAI) to create:

• A newsletter-style summary title and blurb.
• Two short summary blurbs for video overlay texts.
• Two creative image prompts for Leonardo.ai.

This frees you from manually crafting marketing-friendly text.

Step 10: Generate Images via Leonardo.ai

Through the REST HTTP Request node “Leo – Generate Image,” send prompt text to Leonardo.ai with parameters like image size and style preset.

Use a wait node for 30 seconds to allow image generation completion.

Step 11: Retrieve Images from Leonardo.ai

Use “Leo – Get imageId” (HTTP Request node) to query generation results and get final image URLs.

Step 12: Create Videos on RunwayML

Send the generated image URLs with additional parameters via HTTP Request node “Runway – Create Video” to generate video clips with AI-driven motion.

Add a Wait node (3 minutes) for video processing.

Query video URLs with “Runway – Get Video” HTTP node after wait.

Step 13: Improve Secondary Image and Video

Repeat image prompt improvement, generation, and video creation steps with “Leo – Improve Prompt2,” “Leo – Generate Image2,” “Runway – Create Video2,” and corresponding Wait/Query nodes to produce a second video asset.

Step 14: Stitch Final Video with Creatomate

Use the Creatomate HTTP Request node to combine video assets, images, subtitles, and AI-generated voiceovers into a single polished video.

Set template ID and modify elements using data references from previous nodes.

Add Wait node and final query to confirm rendering completion.

Step 15: Upload and Social Share

Upload final videos to Minio, Dropbox, Google Drive, and Microsoft OneDrive nodes for content backup and distribution.

Use YouTube, Twitter (X), Instagram, and LinkedIn nodes to post videos to social channels automatically.

Customizations ✏️

1. Change Data Source: Swap the Hacker News node for any RSS or news API you prefer by replacing the node and adjusting URL fetching HTTP Request accordingly.
2. Adjust Summary Lengths: Edit prompts in the Article Prep (OpenAI) node to customize summary or blurb lengths to fit various social media format requirements.
3. Change Image Style: In the Leo – Generate Image nodes, alter parameters like presetStyle to match your brand aesthetic or seasonal themes.
4. Adapt Video Models: Modify the RunwayML model parameter to experiment with different video creation engines for varied visual effects.
5. Add New Platforms: Insert new nodes for TikTok, Facebook, or LinkedIn API to expand reach, linking video URLs from Creatomate outputs.

Troubleshooting 🔧

Problem: “OpenAI node returns incomplete summary or no output.”
Cause: Insufficient prompt detail or rate limit exceeded.
Solution: Ensure prompt is clear, retry after cooldown, check API quota in OpenAI dashboard.

Problem: “Leonardo.ai image generation fails or returns errors.”
Cause: API key expired or incorrect prompt format.
Solution: Verify API credentials in n8n, review prompt length and content formatting exactly as required by Leonardo.ai.

Problem: “RunwayML video creation stuck in processing.”
Cause: API service downtime or wrong model ID.
Solution: Check RunwayML system status, confirm latest API versions, update model ID in HTTP Request node parameters.

Pre-Production Checklist ✅

• Verify all API credentials for OpenAI, Leonardo.ai, RunwayML, Creatomate, and cloud storage connected correctly.
• Test Manual Trigger and check each node output individually in n8n debug to confirm data flow and no errors.
• Validate HTTP request URLs dynamically pull article content and images as expected.
• Confirm OpenAI analysis filters only relevant AI articles.
• Ensure wait nodes are set appropriately for asynchronous API generations.
• Backup cloud storage destinations configured for fail-safe asset uploads.

Deployment Guide

When ready, activate the workflow in your n8n instance. Use the manual trigger to run test batches before scheduling.

For scheduled runs, add a Cron Trigger at the start for regular article processing.

Monitor execution logs via n8n interface or external monitoring tools for any failure points.

Extend with alert nodes notifying you of failure or completion for operational confidence.

FAQs

Q: Can I replace Hacker News with other news sources?
A: Yes, just swap the data source node and adjust subsequent HTTP requests to fetch article content accordingly.

Q: Does this workflow consume high API credits?
A: It can, depending on volume. Limit the number of articles or refine summary lengths to reduce usage.

Q: Is my data secure in this automation?
A: Yes, especially if self-hosted. All API keys are kept secure in n8n credentials with encrypted storage.

Q: Can it handle scaling to hundreds of articles?
A: Yes, with batch limits and rate-limit respectful wait nodes built in.

Conclusion

By building this custom n8n workflow, you’ve automated the full pipeline from Hacker News article fetching to polished AI-generated videos ready for multi-platform sharing. This saves you hours daily, reduces manual errors, and elevates your content quality with cutting-edge AI tools like OpenAI, Leonardo.ai, RunwayML, and Creatomate.

Next, consider expanding this system to other news sources, adding more social platform integrations, or incorporating real-time video analytics for optimization.

You are now equipped to supercharge your content creation with AI automation — enjoy your new, highly efficient workflow!