AI Agent workflows
518 results — all source-linked n8n references
Grade system prompts in Google Sheets with a Gemini LLM judge
Quick overview This workflow manually grades a system prompt by running it against a sample user task with Google Gemini, then has a second Gemini judge score the prompt (1–10) and write the grade and reason back to Google Sheets. How it works Starts when you run the workflow manually. Reads the prompt list from a Google Sheets spreadsheet and keeps only the last row as the system prompt under test. Sets a fixed sample customer-support task and sends it to a Google Gemini-powered agent using the sheet prompt as its system instructions. Sends the original prompt, the sample task, and the agent’s response to a second Google Gemini judge that returns a structured grade and short justification. Updates the same Google Sheets row with the returned grade and reason. Setup Connect a Google Sheets OAuth2 credential with access to the target spreadsheet. Connect a Google AI Studio (Gemini) credential for the candidate model, grader model, and parser-fixer model. Ensure your sheet has columns for at least prompt, grade, and reason, and that the workflow can match rows using the row_number field. Update the spreadsheet ID/sheet and edit the sample task text if you want to test prompts against
Book, reschedule and confirm WhatsApp appointments with Claude and Sheets
Quick overview This workflow receives WhatsApp messages via a Meta webhook, uses Anthropic Claude to detect booking intents and extract appointment details, writes updates to Google Sheets, and sends a confirmation or error reply back to the customer on WhatsApp. How it works Receives incoming WhatsApp messages through a Meta webhook endpoint and normalizes key fields like sender phone number, display name, message text, and timestamps. Filters out empty messages and sends the text to Anthropic Claude to classify the intent (book, reschedule, confirm, or unknown) and extract structured appointment details. Validates the request based on intent, including required fields and simple rules like weekday-only appointments and business hours. Appends a new row to the Google Sheets Appointments tab for valid bookings, or updates the appointment status/date/time in Google Sheets for reschedules and confirmations. Builds a personalized WhatsApp reply message describing the result (success, required missing details, or supported actions). Sends the reply back to the customer via the WhatsApp Cloud API and appends the interaction details to the Google Sheets ActivityLog tab. Setup Create a Me
Detect competitor launch signals from ads with OpenAI and Google Sheets
Quick Overview This workflow runs weekly to scan competitor ad copy from Meta, Google Ads, and LinkedIn via Adyntel, detects newly emerging terms, uses OpenAI to assess launch signals, and records results in Google Sheets while sending a Slack alert when a launch signal is detected. How it works Runs every week on a schedule (Monday) to start the competitor monitoring cycle. Reads competitors marked as Pending from Google Sheets and processes them one by one. For each competitor domain, pulls current ads from Meta, Google Ads, and LinkedIn using Adyntel and combines the results. Extracts and deduplicates ad titles and body copy, then loads the competitor’s previously known terms from Google Sheets. Extracts candidate new terms from this week’s ad copy, filters for terms that recur across multiple ads, and decides whether there are meaningful new terms to analyze. If meaningful new terms exist, sends the context to OpenAI to classify the terms and determine whether a product launch signal is present, then appends the signal record and the newly classified terms to Google Sheets. If no meaningful new terms exist, appends baseline terms from this week’s ad titles to Google Sheets to p
Extract ecommerce product data with Google Sheets, ScrapingBee and Gemini
Quick Overview This workflow manually runs to read a list of webpage URLs from Google Sheets, scrape each page with ScrapingBee, and use Google Gemini to extract structured product data from screenshots with an HTML fallback, then append the results back into a Google Sheets sheet. How it works Runs when you manually trigger the workflow. Reads the list of URLs to scrape from a Google Sheets spreadsheet. Fetches a full-page screenshot for each URL using the ScrapingBee API. Sends the screenshot (and the URL for context) to a Google Gemini model to extract product details into a structured JSON format, calling a ScrapingBee HTML fetch tool when screenshot extraction is incomplete. Converts any fetched HTML to Markdown and returns it to the Gemini agent to complete the extraction. Splits the extracted product array into individual items and appends them as new rows in the Google Sheets “Results” sheet. Setup Create a Google Sheets service account connection in n8n and set the target spreadsheet and the “List of URLs” and “Results” sheet selections. Add a Google Gemini (PaLM) API credential in n8n and ensure the selected model (gemini-1.5-pro-latest) is available for your account. Add
Analyze skill gaps for target jobs with BrowserAct, LinkedIn and GPT-4o-mini
Quick Overview This workflow collects a student’s skills and target role via an n8n Form, scrapes top LinkedIn job postings with BrowserAct, and uses OpenAI GPT-4o-mini to generate a skill gap report with courses and a 90-day plan, saving results to Google Sheets and emailing them via Gmail. How it works Receives a student submission from an n8n Form with name, email, target role, current skills, experience level, location, and weekly learning hours. Validates the inputs, converts the skills list into an array, and builds a LinkedIn job search URL for the chosen role and location. Uses BrowserAct to open the LinkedIn job search in a real browser and extract text from up to 20 job postings. Cleans and normalizes the scraped job text and combines it with the student profile data. Sends the student profile and job-posting text to OpenAI GPT-4o-mini to extract in-demand skills, identify missing skills with priority, recommend three free courses, and generate a 13-week learning plan. Parses the model output into labeled sections, calculates matched and missing skill counts, and appends the full report as a new row in Google Sheets. Sends a formatted HTML email via Gmail to the student w
CQualify LinkedIn leads and draft multi-step outreach with Airtable, Apify and Claude
Quick overview This workflow pulls unprocessed leads from Airtable, enriches them with LinkedIn profile, company, and post data via Apify, optionally scrapes the company website with Firecrawl, then uses Anthropic Claude to qualify leads and draft multi-step LinkedIn outreach messages, saving results back to Airtable. How it works Runs when you manually execute the workflow. Reads qualification and writing prompts from Airtable, then searches Airtable for leads where the “Réponse” and “Match” fields are empty. Iterates through each lead and uses Apify to fetch LinkedIn profile details, marking the Airtable record as “Likedin pas dispo” when the profile cannot be retrieved. For available profiles, uses Apify to fetch company details when a company LinkedIn URL exists and, if a website is available, uses Firecrawl to scrape the homepage and clean the returned markdown. Uses Apify to fetch recent LinkedIn posts for the lead and selects a relevant recent personal post URL to include in the enrichment dataset. Sends the consolidated lead, company, homepage, and post data to Anthropic Claude to decide whether the lead matches the qualification criteria, marking non-matching leads as “Fal
Score leads and send WhatsApp proposals with Gemini, HighLevel and Redis
Quick Overview This workflow captures new leads via webhook, enriches them with Apify-powered Google AI search, scores them using Google Gemini, and updates HighLevel CRM while sending WhatsApp template messages; it also handles inbound WhatsApp support with a Gemini-driven agent that uses Redis memory, updates HighLevel, and notifies a team via Telegram. How it works Receives new lead details through an HTTP webhook. Normalizes the lead fields and uses Apify (Google AI Mode search) to fetch a public professional background overview for the lead. Sends the enriched lead context to Google Gemini to produce a structured lead score, rationale, recommendation, and next steps. Routes the lead based on the score threshold and upserts the contact in HighLevel with the AI scoring details stored in custom fields. Creates a HighLevel opportunity, sends a WhatsApp template message to the lead, and updates the opportunity’s pipeline/stage fields. Triggers on incoming WhatsApp messages, looks up the sender in HighLevel, fetches related opportunities, and hands the conversation to a Google Gemini agent with Redis chat history. Uses the agent’s decisions to save requirement notes and update/close
Draft and revise technical articles with OpenRouter in a content pipeline
Quick Overview This subworkflow is triggered by a parent content pipeline, uses OpenRouter (via an AI Writer Agent) to draft or revise a technical article from a brief, research notes, and optional reviewer feedback, and returns the updated pipeline state with the current draft and word count. How it works Receives input from a parent workflow containing the brief, research notes, and optionally a previous draft plus reviewer feedback. Sends the brief, research JSON, and any revision context to an AI Writer Agent backed by OpenRouter to generate a fresh draft or revise the prior draft. Parses the agent’s JSON response, extracts the draft text and word count (or calculates it if missing), and flags whether the output is a revision. Merges the extracted fields back into the original input and returns the updated state (including currentDraft) to the parent workflow. Setup Add an OpenRouter API credential and select it for the OpenRouter chat model used by the Writer Agent. Ensure the parent workflow passes the expected fields (for example brief, research, and optionally previousDraft and feedbackNotes) when calling this subworkflow. Update the Writer Agent system prompt if you need a
Score and critique content drafts with OpenRouter and LangChain
Quick Overview This subworkflow is called by a parent n8n pipeline to review a content draft against a provided brief using an OpenRouter chat model, returning structured scores, issues, and revision notes the parent can use to decide whether to iterate or accept the draft. How it works Receives the full pipeline state from a parent workflow execution, including the brief and current draft text. Sends the brief and draft to a LangChain Agent configured as an editorial “LLM-as-a-Judge” using an OpenRouter chat model. Produces a JSON review containing 1–10 scores for accuracy, tone, completeness, and clarity, plus an averaged overall score, issues list, and actionable revision notes. Parses the model output, falls back to a zero-score error review if parsing fails, and merges the review back into the original input object. Returns the enriched payload (original fields plus a review object) to the parent workflow for threshold-based decision-making. Setup Add an OpenRouter API credential and select the model you want to use in the OpenRouter chat model node. Ensure the parent workflow passes the expected fields (at minimum brief and currentDraft) when calling this subworkflow. Save th
Reconcile Google Drive invoice PDFs with OpenAI, Sheets, Slack and Gmail
How it works Watches a Google Drive folder and triggers automatically when a new invoice PDF is uploaded Extracts raw text from the PDF and passes each invoice through an OpenAI powered agent that returns structured fields like vendor, amount, due date, and line items Fans the enriched data to three parallel outputs: a Google Sheets log, a Slack alert for overdue invoices, and a weekly summary email to Finance via Gmail Set up steps Connect Google Drive, Google Sheets, Gmail, Slack, and OpenAI credentials (~15 minutes total) Works best with text based PDFs; scanned invoices need OCR upstream of the extraction step Keep full setup details and prompt tuning tips in the sticky notes inside the workflow
Generate client-ready workflow diagrams from sales call transcripts with Claude
Who it's for This workflow is for automation agencies and n8n consultants who conduct discovery calls with prospects and want to instantly turn conversation transcripts into polished, client-ready workflow diagrams — without manual canvas work. How it works A consultant pastes a sales call transcript into a form trigger. A first AI agent (Claude) parses the transcript and extracts structured workflow briefs — one per identified business process — with titles, tools, logic, and step-by-step descriptions. A code node splits the output into individual workflow items, one per identified process. A second AI agent generates a complete n8n workflow JSON for each brief, using French business-oriented node names and proper canvas positioning. A code node validates and parses the generated JSON to ensure structural integrity. A third AI agent clusters nodes into logical business groups and generates canvas layout coordinates and sticky note annotations. A fourth AI agent renames all nodes with clear, descriptive French labels following a consistent naming convention. All renames and layout changes are applied, and the finalized workflow is published directly to an n8n instance via the API.
Analyze stocks via Telegram with GPT‑4, TwelveData, NewsAPI and chart images
AI-Powered Stock Analysis Agent is a Telegram-based AI workflow it transforms a simple stock name into a complete trading analysis report. Users send a stock/company name through Telegram, and the workflow automatically: Converts the company name into a stock symbol Fetches multi-timeframe market data Collects recent stock news Performs AI-powered sentiment analysis Generates a chart image Creates a professional BUY / SELL / HOLD report Sends the final analysis back to Telegram Node-by-Node Explanation 1. Telegram Stock Request Trigger What This Node Does This is the entry point of the workflow. It listens for incoming Telegram messages and starts the workflow automatically whenever a user sends a stock/company name. Example Apple Tesla Nvidia 2. Convert to Stock Symbol What This Node Does Uses OpenAI to convert a company name into a valid stock ticker symbol. Example | User Input | AI Output | |---|---| | Apple | AAPL | | Tesla | TSLA | | Microsoft | MSFT | Why This Node Matters Users do not need to know the exact stock ticker symbol. The AI intelligently maps company names into valid trading symbols. 3. Fetch 4H Market Data What This Node Does Fetches 4-hour interval market data
WTurn videos into Twitter threads, LinkedIn posts, and blogs with WayinVideo and GPT-4o-mini
Description Submit any video URL — podcast, webinar, interview, or YouTube video — along with the title, creator name, niche, target audience, and tone and the workflow transcribes the full recording automatically. WayinVideo's Transcription API returns a speaker-labeled, timestamped transcript which is sent to GPT-4o-mini in one call to produce all three content pieces simultaneously: a 10-tweet Twitter thread, a 200–300 word LinkedIn post, and a 600–800 word SEO blog post with HTML structure, meta description, and focus keyword. Word counts are tracked for all three pieces and everything is saved as one row in Google Sheets with Status set to Draft. Built for content marketers, solo creators, agencies, and social media managers who want to repurpose every video into three platform-specific pieces without writing anything manually. What This Workflow Does Transcribes the full video with speaker labels and timestamps** — WayinVideo's Transcription API returns every word spoken in the video with the speaker name and timestamp so GPT has full context to work from Generates all three content pieces in one GPT call** — The Twitter thread, LinkedIn post, and blog post are all produced t
WTurn webinars into a 5-day email course using WayinVideo, GPT-4o-mini, and Google Sheets
Description Submit any webinar recording URL using a simple form along with your course name, target audience, brand name, and preferred email tone and the workflow builds a complete 5-day drip email course automatically. WayinVideo AI Clipping extracts the top 5 most engaging teaching moments from the recording, ranked by engagement score so the best content becomes Day 1. GPT-4o-mini then writes a complete email lesson for each clip — with subject line, preview text, email body, key takeaway, CTA text, and a CTA URL placeholder — where Day 1 is welcoming, the final day wraps up, and middle days flow naturally between them. All 5 emails are saved to Google Sheets as a ready-to-use drip course with Status set to Draft. Built for course creators, coaches, educators, and marketing teams who want to turn any webinar into a ready-to-send email sequence without writing a single word manually. What This Workflow Does Extracts the top 5 teaching moments by engagement score** — WayinVideo AI Clipping ranks moments by engagement and the workflow sorts them highest first so your best content always becomes Day 1 Assigns each clip to a day number automatically** — Clips are sorted and labeled
Automate employee onboarding with Groq, Gmail, Slack and Google Workspace
How it works New hire submits an onboarding form with their details Groq AI generates personalised welcome content, Slack messages, and IT access request Welcome email sent to the new hire via Gmail Automated emails sent to HR, the manager, and IT team Company Policy and Employee Handbook PDFs downloaded from Drive and attached Slack messages posted to #general and #hr-team channels Onboarding record appended to Google Sheets tracker Set up steps Estimated time: ~15 minutes Connect Groq API, Gmail, Slack, Google Drive, and Google Sheets credentials Upload your Policy and Handbook PDFs to Drive and link them in the Download nodes Select your Slack channels in the two Slack nodes Select your onboarding Google Sheet in the Append row node Activate the workflow and share the form link with new hires
WTurn TikTok videos into Facebook posts using WayinVideo and GPT-4o-mini
Description Paste any TikTok, YouTube, or short-form video URL into a simple form along with your brand name, target audience, and preferred tone and the workflow automatically transcribes the content and writes a ready-to-publish Facebook caption. GPT-4o-mini follows ten Facebook engagement rules to write a hook, short paragraphs, a call to action, and up to 8 hashtags — then posts the caption directly to your Facebook Page feed via the Graph API. Every post is logged to Google Sheets with the Post ID, Post URL, character count, and Posted status. Built for social media managers, content creators, and brands who produce short-form video and want Facebook posts published automatically without rewriting captions manually. What This Workflow Does Transcribes any video into plain text** — WayinVideo processes the video and returns a clean spoken text version that GPT uses as the source for caption writing Writes a Facebook caption following 10 engagement rules** — GPT-4o-mini produces a hook, short paragraphs, a tone-matched body, a CTA, optional video link, and up to 8 hashtags Optionally includes the video link in the post** — If you choose Yes at form submission, the video URL is a
WConvert podcasts to LinkedIn articles with WayinVideo and GPT-4o-mini
Description Submit your podcast or video URL using a simple form and the workflow automatically transcribes the full episode with speaker labels and timestamps. GPT-4o-mini then reads the transcript and writes a complete LinkedIn newsletter article — headline, 400–600 word body, five key takeaways, and hashtags — following seven LinkedIn readability rules. The finished draft is saved to Google Sheets with a Status of "Draft" so you can review and publish when ready. Built for podcast hosts, content marketers, and personal brand builders who want to turn every episode into a LinkedIn article without writing from scratch. What This Workflow Does Transcribes your podcast with speaker labels** — WayinVideo processes the full episode and returns each segment attributed to the correct speaker with timestamps Polls automatically until transcription is complete** — Waits 60 seconds then checks the status repeatedly every 30 seconds until the transcript is ready, with no manual monitoring needed Formats the transcript for AI analysis** — Structures every segment as speaker name with timestamp so GPT understands who said what and when Writes a complete LinkedIn article** — GPT-4o-mini produc
Generate AI-powered pro vs con debates with GPT-4o and Google Sheets tracking
This workflow automatically processes incoming debate topic requests, generates structured, evidence-backed Pro vs Con debates with balanced arguments, rebuttals, and final verdicts using AI, delivers them via email, and logs everything in Google Sheets. Who’s it for • Students and debaters preparing arguments • Content creators and podcasters • Researchers and analysts needing balanced views • Decision-makers and AI enthusiasts exploring topics How it works / What it does Captures new debate requests (webhook or scheduled poll) Validates the topic and prepares context AI generates full structured debate (Pro/Con arguments with evidence, rebuttals, verdict) Formats professional markdown output Emails the complete debate to the requester Logs topic, verdict summary, and generation details in Google Sheets How to set up Import this workflow Set up credentials (Webhook, Google Sheets, OpenAI/Anthropic) Update default instructions and email sender Activate workflow Requirements • Webhook endpoint for topic submissions • Google Sheets • OpenAI / Anthropic / Grok API • Clear debate topic input How to customize the workflow • Modify AI prompt for depth, tone, or format in the AI node • Ad
Audit multi-page CRO with GPT-4o-mini, Google Sheets, and Gmail alerts
Description Submit up to 3 landing page URLs using a simple form and the workflow audits each one automatically. For every page, GPT-4o-mini scrapes the content, scores it across five dimensions, identifies the top conversion issue, and generates five specific page fixes. Every result is logged to Google Sheets and pages that score below your alert threshold trigger an automatic Gmail alert with the full breakdown. Built for marketing managers, agency owners, and growth teams who need structured, repeatable conversion audits without spending hours reviewing pages manually. What This Workflow Does Scrapes and cleans each page automatically** — Fetches the raw HTML for every URL submitted, strips all scripts and styles, and prepares clean text for AI analysis Scores pages across five CRO dimensions** — GPT-4o-mini rates each page on overall conversion potential, CTA strength, trust score, and message clarity on a 1–10 scale Generates a letter grade per page** — Assigns A+ through F based on the overall score so you can rank pages at a glance Produces five page-specific fixes** — Every fix references actual content from the page and is scoped to be doable within 1–2 days — no generic
Optimize blog and product pages for Google AI Overviews with GPT-4 and Sheets
This workflow automatically monitors blog posts or product pages for new/updated content, analyzes Google AI Overviews to identify optimization gaps, auto-generates optimized titles, summaries, and schema markup, and applies these improvements to enhance visibility in AI search results. Who's it for • SEO managers optimizing for AI-driven search results • Content teams publishing 5+ articles per week • E-commerce teams managing product page visibility • Digital marketers tracking AI Overview presence How it works / What it does Detects new or updated blog posts / product pages via webhook or schedule Fetches raw content and existing metadata Queries Google Search API to analyze AI Overview coverage gaps AI generates optimized title, meta summary, and JSON-LD schema markup Pushes improvements back to CMS or target endpoint Logs all changes and scores to Google Sheet tracker How to set up Import this workflow Set up credentials (Google Search API, OpenAI/Anthropic, CMS/HTTP, Google Sheets) Update your site URL and content preferences Activate workflow Requirements • Webhook endpoint or CMS trigger • Google Custom Search API key • OpenAI / Anthropic / Grok API • CMS REST API or target
SAnswer portfolio questions with live profile data, Groq and chat memory
Build a conversational AI chatbot for your personal portfolio that answers questions about your experience, skills, projects, and contact info — powered by Groq and live profile data fetched from a JSON endpoint. How it works A visitor sends a message via the public chat widget The workflow fetches your latest profile data from a hosted JSON URL Profile data + chat input are prepared and passed to the AI agent The Groq-powered agent answers questions using only your profile data Chat memory retains the last 10 messages for context-aware replies Setup Replace the profile JSON URL in Get Profile with your own hosted endpoint Connect your Groq API credential Update the system prompt in Chat Agent with your name Activate and embed the chat widget on your portfolio site Requirements Groq API key (free tier works) A publicly accessible JSON file containing your profile data
Route and triage support tickets with Claude Sonnet and your CRM
Overview This workflow automates customer support ticket processing using AI-powered analysis, classification, and intelligent routing. It processes incoming tickets from email or webhook, translates messages when needed, analyzes sentiment and urgency, and routes tickets to auto-reply or escalation flows. The system also updates CRM platforms and logs observability metrics for monitoring. This enables faster response times, improved customer experience, and scalable support operations. How It Works Input Sources Receives tickets via: IMAP Email Trigger Webhook endpoint Workflow Configuration Defines: CRM/Helpdesk API endpoint Escalation webhook URL Observability logging endpoint Data Cleaning & Normalization Extracts and cleans HTML content Normalizes ticket data: Ticket ID User email Message content Timestamp Source channel Language Detection & Translation Detects the original language Translates message into English if needed Returns confidence score AI Support Intelligence Classifies ticket into: Sentiment (positive/neutral/negative) Urgency (low → critical) Category (billing, bug, technical, etc.) Generates: Short summary Churn risk score (0–1) Recommended action path Decision
SOrchestrate credit onboarding checks with GPT-4o, Airtable, Gmail and Slack
How It Works This workflow automates credit operations onboarding by running KYC verification, credit bureau checks, identity validation, and sanctions screening through a single AI-powered agent. Built for credit operations teams, compliance officers, and fintech platforms, it eliminates manual eligibility reviews that are slow and error-prone. Triggered via webhook, the Credit Operations Agent orchestrates all verification tools simultaneously, then routes customers by eligibility status, eligible, ineligible, pending documentation, or compliance escalation. Each path prepares structured data stored in Airtable, triggers appropriate follow-up actions (email, Slack alerts), and logs a full audit trail. A final formatted response is returned to the originating system, closing the loop end-to-end with no manual handoffs. Setup Steps Set webhook URL and connect Credit Operations webhook node to your intake system. Add OpenAI API key to the OpenAI Chat Model node. Configure KYC, Credit Bureau, Identity, and Sanctions tool credentials. Add Gmail OAuth2 and Slack bot token for notification nodes. Connect Airtable API key; set base/table IDs for eligible and ineligible customer stores. P
Monitor and optimize carbon emissions for ESG reporting with GPT-4o, Slack and Sheets
How It Works This workflow automates end-to-end carbon emissions monitoring, strategy optimisation, and ESG reporting using a multi-agent AI supervisor architecture in n8n. Designed for sustainability managers, ESG teams, and operations leads, it eliminates the manual effort of tracking emissions, evaluating reduction strategies, and producing compliance reports. Data enters via scheduled pulls and real-time webhooks, then merges into a unified feed processed by a Carbon Supervisor Agent. Sub-agents handle monitoring, optimisation, policy enforcement, and ESG reporting. Approved strategies are auto-executed or routed for human sign-off. Outputs are consolidated and pushed to Slack, Google Sheets, and email, keeping all stakeholders informed. The workflow closes the loop from raw sensor data to actionable ESG dashboards with minimal human intervention. Setup Steps Connect scheduled trigger and webhook nodes to your emissions data sources. Add credentials for Slack (bot token), Gmail (OAuth2), and Google Sheets (service account). Configure the Carbon Supervisor Agent with your preferred LLM (OpenAI or compatible). Set approval thresholds in the Check Approval Required node. Map Googl