Evaluations metric: answer similarity
This n8n template demonstrates how to calculate the evaluation metric "Similarity" which in this scenario, measures the consistency of the agent. The scoring approach is adapted from the open-source evaluations project RAGAS and you can see the source here https://github.com/explodinggradients/ragas/blob/main/ragas/src/ragas/metrics/answersimilarity.py How it works This evaluation works best where questions are close-ended or about facts where the answer can have little to no deviation. For our scoring, we generate embeddings for both the AI's response and ground truth and calculate the cosine similarity between them. A high score indicates LLM consistency with expected results whereas a low score could signal model hallucination. Requirements n8n version 1.94+ Check out this Google Sheet for a sample data https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit?usp=sharing
By JimleukAuto-post YouTube Shorts from Google Drive with OpenRouter AI titles & descriptions
How it works This workflow automatically uploads YouTube Shorts from a Google Drive folder. It picks one video at each run, generates a YouTube-optimized title, description, and hashtags using an AI model, uploads the video through YouTube’s resumable upload API, and finally moves the processed video to a “Posted” folder. Key steps Run on a schedule (daily/hourly/custom CRON) Fetch one video from a selected Google Drive folder Generate title + description + hashtags using an LLM Prepare YouTube metadata Upload the video via resumable upload Move the file to a “posted” folder to avoid duplicates Setup instructions Connect your Google Drive credentials Connect your YouTube OAuth2 credentials Update the Drive folder IDs (input + posted folders) Edit the “Set variables” node to change store name, country, coupon code, and tone Review the prompt in the AI node if you want to customize content style All technical details are documented inside the sticky notes within the workflow.
By osama godaScan URLs with urlscan.io and send results via Gmail
Overview Receive a URL via Webhook, submit it to urlscan.io, wait ~30 seconds for artifacts (e.g., screenshot), then email a clean summary with links to the result page, screenshot, and API JSON. What this template does Ingests a URL from a POST request. Submits the URL to urlscan.io and captures the scan UUID. Waits 30s to give urlscan time to generate the screenshot and result artifacts. Sends a formatted HTML email via Gmail with all relevant links. Nodes used Webhook (POST /urlscan) urlscan.io → Perform a scan Wait (30 seconds; configurable) Gmail → Send a message Input json { "url": "https://example.com" }
By Calistus Christian