Video summary
Tutorial de n8n para Principiantes 2026: Crea Agentes de IA Paso a Paso
Main summary
Key takeaways
Summary
The video is a beginner-friendly, step-by-step tutorial for building an automated AI-agent workflow in n8n. The workflow creates a daily email newsletter with weather, news, and trivia, avoids repeating previous topics, and records each report in Google Sheets.
Tutorial and Key Features
1. Create an n8n workflow
- Start with a chat-message trigger and add an AI Agent node.
- Connect a large language model (LLM) to serve as the agent’s reasoning engine. The tutorial uses OpenAI, though n8n supports other providers.
- OpenAI API usage is billed by tokens and is separate from a ChatGPT subscription. Store the API key securely.
2. Give the agent memory and tools
- Add simple memory so the agent can use recent interactions as context.
- Add Wikipedia for facts and SerpApi for web searches. SerpApi offers a limited number of free calls.
- Test the agent by asking it to find current news and an animal fact.
3. Automate the workflow
- Replace the chat trigger with a Schedule Trigger set to run daily at 8 a.m.
- Prompt the agent to create a daily report that includes San Francisco weather, positive news, and a Wikipedia fact.
- When there is no chat trigger, configure memory with a session ID. Otherwise, the workflow may fail because no session ID is available.
4. Send a formatted report
- Add OpenWeatherMap as a tool and configure an API key and city.
- Add Gmail to send the agent’s output to an inbox. Connecting a Google account may require OAuth configuration, particularly with self-hosted n8n.
- Convert the agent’s Markdown output to HTML, then use a separate OpenAI step to improve the email’s appearance with inline CSS.
5. Prevent duplicate content
- Create a Google Sheet with columns for the date, news, and fun facts.
- Use a separate AI step to extract the report’s items and add them as a new row.
- Before generating each report, use another AI step with Google Sheets access to review past entries and summarize topics to avoid.
- Pass that summary into the main agent’s prompt. The tutorial recommends dividing tasks across AI steps rather than overloading one agent with unrelated jobs.
6. Test, activate, and maintain
- Run the entire workflow before enabling it, inspect node outputs, and save the workflow.
- Once activated, the agent sends a daily report and records its contents to help prevent repetition.
- The presenter recommends giving clear instructions, limiting tools to those needed, using prompt templates and variables, and expecting some trial and error. n8n’s AI assistance and node documentation can help with troubleshooting.
Templates and Hosting
- n8n’s template library offers ready-made workflows that can be copied into a blank workflow and customized with credentials and prompts.
- The video compares n8n Cloud with self-hosting on a VPS:
- Cloud handles hosting and maintenance.
- Self-hosting offers more control, privacy, and potentially lower cost, but requires responsibility for setup, updates, security, and uptime.
- Running n8n on a personal computer means workflows stop when that computer is off.
Main Speaker and Source
Metics Media | Español — the video’s narrator and tutorial presenter.
Rate this summary
Your feedback will help improve summaries.
Improve this summary
Reprocess with a stronger model when the summary feels incomplete or inaccurate.
Translate summary in another language
Ask questions to this video
Chat for follow-up questions, clarifications, and source-backed answers.