Extract From File (PDF)
Tengo problemas con el nodo Extract From File - PDF. El error es el siguiente: Invalid Root reference. Este es el workflow ¿Alguna sugerencia? 3 posts - 3 participants Read full topic

Posts from every feed QUEST has found across the web — blogs, news, podcasts, channels and more.
Tengo problemas con el nodo Extract From File - PDF. El error es el siguiente: Invalid Root reference. Este es el workflow ¿Alguna sugerencia? 3 posts - 3 participants Read full topic

Describe the problem/error/question I have several workflows that are being triggered 4-5 times on a single scheduled interval, when I only expect them to be triggered once. There is no record of these executions in the execution history except for the last (most recent) one, which suggests to me this is a problem deep somewhere in the way n8n handles scheduled triggers. I suspect this has something to do with schedueld triggers being updated via MCP, because I never had this issue prior to…
Hey n8n community! I recently finished building an AI-Powered Freelance Opportunity Aggregator with n8n as the orchestration layer. The goal was not just to collect job listings, but to build a complete decision-support pipeline that discovers opportunities from multiple source types, normalizes the data, filters obvious mismatches, evaluates the remaining opportunities with structured LLM output, and delivers prioritized results through Telegram. Architecture Discovery → Normalization →…

I’m working on how to add AI agents to Jira-like workflows, and n8n turned out to be the perfect glue around it. Setup: Jira tickets are synced with an agent tool (I use Sharkly.ai, disclosure: I work on it). The agent does the coding task and writes its result back as a Jira comment, since the sync is bidirectional. n8n handles everything around that: Jira Trigger on comment created / issue updated. IF node: is the comment from the agent account? Slack / Email node: notify the reviewer with…
I’m looking for firsthand experiences from people who hired someone to evaluate an existing document-based assistant, retrieval pipeline, or chatbot. If you’ve purchased this service: • What prompted you to hire an external evaluator? • How much did you pay? A range and currency are fine. • What did the evaluation include, and what did you receive? • What made it worth the money—or disappointing? • Did the findings change a release decision or lead to a specific improvement? • Where did you…
Hi! I build and maintain n8n automations with a real backend behind them, not just drag-and-drop. What I do: n8n workflows end to end: integrations, webhooks, OAuth, retries, idempotency, error handling, cost monitoring. Self-hosted or cloud. AI / LLM steps: OpenAI, Claude, Gemini via API - RAG with pgvector, semantic search, prompt and output validation, call/chat analysis pushed back into CRM. Python when no-code hits a wall: custom nodes, API integrations, web scraping (proxies, anti-bot),…
Hey everyone! Challenge 06: The Support Queue is live! A support desk has 24 tickets. Can your native n8n workflow count the open tickets, then find how many of those are urgent? One personal dataset, two counts, and the third challenge of Season 2. Your existing username and participant key still work, and your Challenge 04 and 05 points stay on the season board. Start the challenge · Get the starter · Questions & Discussion Your timer runs from the first successful personal fetch to the first…
Hey everyone! Challenge 06: The Support Queue is live! This is the third challenge of Season 2. A support desk has 24 tickets. Build a native n8n workflow to count how many are open, then how many of those open tickets are urgent. Both questions use the same personal dataset. Start Challenge 06 · Get the starter workflow · Read the rules Your existing username and participant key still work. Your Challenge 04 and 05 points stay in your Season 2 total. The first successful personal data fetch…
Hi everyone! I’m Illia, an automation engineer based in Ukraine (UTC+3). My hours fully overlap with Europe and with US mornings. What I build in production: Self-hosted n8n on Docker/Proxmox + PostgreSQL: backups, credentials in env vars, error workflows with Telegram alerts LLM agents with Claude / OpenAI: email reply drafting with human approval, lead qualification, booking assistants (Telegram/WhatsApp) Lead sourcing pipelines: SerpAPI → AI classification → personalization → email finding →…
OpenAI is removing gpt-image-1 from the API on 23 October 2026. Its listed replacements are gpt-image-2.5-sunburst or gpt-image-2.5-flare. In n8n, the OpenAI node’s Image → Edit Image operation uses gpt-image-1 as its default model. Any workflow that still has it selected will start failing on that date. Am I affected? Open each workflow that has an OpenAI node set to Image → Edit Image and look at the Model field. If it says GPT Image 1 / gpt-image-1, you need to change it. Edit Image steps…
Describe the problem/error/question Can someone help me understand the different ways to access data from the first branch in the second branch in n8n, other than using the Merge node? Are there any alternative nodes or approaches that can help achieve this, especially when both branches may not always contain data? What is the error message (if any)? Please share your workflow (Select the nodes on your canvas and use the keyboard shortcuts CMD+C/CTRL+C and CMD+V/CTRL+V to copy and paste the…
Trusted Metadata Inputs for the New n8n Agent Builder Hi n8n community, I’m exploring the new n8n Agents / Agent Builder and have encountered a limitation when passing trusted metadata to tools. The problem Currently, I haven’t found a native way to pass metadata directly to a tool independently of the AI agent’s generated arguments. For example, consider a shopping agent with an Order History tool. The agent should determine the user’s intent, but it shouldn’t control the identity or tenant…
Looking to Build a Small BPA Mastermind Group Hey everyone! I’m currently learning Business Process Automation (BPA) using n8n, with the goal of turning it into a service business. I’m looking to connect with 2–3 serious people who are learning or working in n8n, workflow automation, AI automation, or freelancing. The idea is to create a small online mastermind group where we can: Share what we’re building and learn from each other’s experience. Help each other troubleshoot workflows and solve…

Hi everyone! I’m using n8n Cloud with a Neon PostgreSQL database. My connection works successfully with: SSL: Require Ignore SSL Issues: OFF Port: 5432 SSH Tunnel: OFF However, when I execute this query through n8n: SELECT ssl, version, cipher FROM pg_stat_ssl WHERE pid = pg_backend_pid(); It returns: ssl = false, version = null, cipher = null Could this be caused by Neon’s connection proxy? Is the external connection still encrypted with TLS? How can I verify that n8n Cloud is properly…
Hi everyone, I built two free tools for n8n users and would love your feedback. I am the author. Workflow Doctor: paste an exported workflow JSON and it checks for broken connections, missing credentials, open webhooks, hard-coded secrets and expression mistakes. It suggests repairs that you review before exporting. Free n8n workflow doctor | TodoToolsOnline Workflow Builder: a visual editor that imports and exports n8n workflow JSON, with 8 templates and a test mode that simulates the flow.…
Hi everyone, I build Vivere (https://vivere.dev), a heartbeat monitor, and I’ve published a community node for it: n8n-nodes-vivere. The problem The Error Workflow only fires when an execution fails. If a Schedule Trigger never fires (the instance was restarting at that minute, the workflow was deactivated after an import, a trigger quietly died), there’s no execution, so nothing reports it. A run that goes green but processes 0 items looks fine too. This comes up in threads here regularly.…
Some failures never turn an execution red. An upstream form field gets renamed and an IF node starts sending every item down the false branch. Someone duplicates a sheet to tidy it, the team moves to the copy, and the Sheets node keeps writing to the original. A form tool’s webhook gets switched off, so the workflow is never triggered. The executions list shows Succeeded, or shows nothing at all, because nothing failed. A check that counts records catches these, where error alerts cannot: Once…
Hi all! I published two community nodes for pulling YouTube data into workflows without the YouTube API or quotas: YouTube Transcripts (T3 Data): n8n-nodes-t3data-youtube-transcripts Videos, Shorts, playlists or whole channels (@handle) Plain text, timestamped segments, SRT, or AI-ready chunks (overlapping, with a timestamped link per chunk), so you can wire it straight into a vector store or AI Agent Language preference, and human captions preferred over auto-generated ones YouTube Comments…
I’m looking for an opportunity as an n8n Developer / AI Automation Engineer. What’s actually built, not “worked on”: → 36 workflows, 3 production systems, not tutorial projects, systems that run → Multi-agent AI stacks wired into real databases, not a spreadsheet pretending to be a backend → APIs, webhooks, end-to-end business logic. The unglamorous plumbing that decides whether a company’s ops actually work or just look like they do Every system is documented down to the architecture, not a…
Hi everyone, Say you have back-to-back sales calls. On the 2pm call, the prospect asks for examples of your work and how soon you could start. By the time you sit down to write the follow-up that evening, you are not sure what you promised. (The call is made up, but most of us have been there.) This workflow writes it for you. A few minutes after Fireflies finishes the transcript: OpenAI writes a short summary, the questions the prospect asked, the next step and an outcome: go ahead, needs…

Turn 60 AI companies’ job boards into a daily hiring-intelligence feed on Telegram. What it does Every morning this workflow pulls ~700 AI/ML job postings from 60 AI companies (via the AI/ML Jobs Scraper actor on Apify), filters for hot hiring signals, groups them by company, and sends one Telegram message per company. Hot signals = roles open 35+ days (stuck hiring), reposted jobs, or max salary $200k+. How it works (5 nodes) Schedule Trigger — daily 08:00 Run AI Jobs Actor — pulls the full…
*Disclosure: I work on the APIGOTO team. APIGOTO is an independent OpenAI-compatible API gateway and isn’t affiliated with OpenAI, Anthropic or Google. I use it as the example below, but every step works with any OpenAI-compatible endpoint (a self-hosted proxy, LiteLLM, vLLM, LM Studio, etc.).* I’ve seen a few threads asking how to use Claude- or Gemini-family models (or a local model) through n8n’s OpenAI nodes, so here’s what I’ve learned setting it up. ## 1. Set the Base URL on the…
I built a set of four n8n templates around one job: checking whether email addresses are worth sending to, before they go into a sequence, a CRM, or an agent’s hands. The templates Google Sheets lead list - reads rows with no verdict, verifies in batches, writes status / score back. Re-runs only pick up new rows, so it’s safe on a schedule. Clean a CSV before cold email - upload an Apollo / Instantly / HubSpot export, get back three lists: send_safe, review, drop. Greylisted addresses wait and…
Hi everyone, I need a small correction on my n8n Academy account. When I signed up I made a typo in my name, which currently shows as “SImone Basile” with a capital I instead of “Simone Basile”. My Academy profile name needs to be corrected, and my n8n Quickstart certificate, which I earned today, should be reissued with the corrected name. My platform ID is f12ce058e514de9aaafc94613a7b1749. The support team pointed me to this category for the request. Thanks in advance for your help! Simone 3…
1. إعداد العقد الأساسية في n8n (Nodes Setup) قم بإنشاء Workflow جديد في n8n وأضف العقد التالية: Webhook Node (أو Chat Trigger): لاستقبال رسائل العملاء (سواء من واتساب، تيليجرام، أو واجهة دردشة موقع الويب). AI Agent Node (Advanced AI): وهو القلب النابض الذي سيقوم بالدور المطلوب. Model Node (OpenAI / Anthropic): يُربط بداخل الـ AI Agent لتزويد النموذج بالذكاء الاصطناعي، مع تمرير System Prompt يحتوي على التعليمات والسياسات الخاصة بالعيادة (الترحيب، عدم إعطاء تشخيصات طبية، جمع البيانات). Google…
Hi all, sharing this in case it saves someone a broken workflow. Reddit is switching off its RSS feeds on 13 November 2026. Workflows that read a subreddit or a Reddit search through the RSS Read node will stop returning items that day, usually without an error you would notice. Disclosure first: I’m the founder of prolevelmax.com, which is the data source used below. If you already have Reddit API access set up, the built-in Reddit node is the cheaper route and you don’t need any of this. The…

Hi all, I’m Jessy, and I build n8n workflows for small UK businesses. This one is free in the template library. I’m posting it because the interesting part comes after the extraction. What it does It watches one Gmail label for supplier invoices. It pulls the PDF text and has OpenAI return a fixed set of fields. Then it checks those fields in plain code before writing a single row. Template: Extract and validate Gmail invoice PDFs into Google Sheets with OpenAI | n8n workflow template How it…
Hey everyone If you’re self-hosting n8n Community Edition, you already know the pain: no built-in version control. Enterprise has Source Control at €667/mo, but for Community users, there’s nothing. I got tired of manually exporting workflow JSON, committing it, and fixing credentials by hand when promoting from staging to production. So I built n8n-sync — a CLI that gives you Git-friendly version control for your n8n workflows. What it does: n8n-sync pull — exports all workflows to normalized…

I built a small template that turns a web page into chunks you can embed directly, using only built-in nodes (HTTP Request + Split Out). Why: fetching a page with HTTP Request and passing it on gives the model mostly navigation, scripts and cookie banners, and splitting by fixed length cuts sections in half. The template asks an API for the page split along its headings instead. Each chunk comes with a stable UUID (same page + position = same id, so a re-run overwrites instead of duplicating),…
Announcements The idea is: Add a Verify action to the Crypto node, next to the existing Sign action. Verify takes the original value, the signature, the algorithm, and the encoding, and checks them with a PEM public key stored on the Crypto credential. The node writes a boolean: true when the signature matches, false when the value was changed or the signature does not belong to that key. My use case: A webhook or API callback arrives with a signature made by a private key (RSA-SHA256, base64…