# n8n AI Agent Tools: How the Agent Gets Things Done > n8n AI agent tools explained: Google Sheets Tool and Gmail Tool vs regular nodes, naming tools in the system message, and limiting a tool to drafts. Source: https://learn.ynteractive.com/content/blog/n8n-ai-agent-tools · Updated 2026-11-10 · Free from Stepthrough (https://learn.ynteractive.com) Quick answer In n8n, a tool is a node you attach under an [AI Agent](https://learn.ynteractive.com/content/glossary/ai-agent)'s Tool slot, such as the Google Sheets Tool or the Gmail Tool, and the agent decides when to use it based on its instructions. A regular node, by contrast, runs every time in a fixed order. This guide is for anyone building their first n8n agent, and it uses the free [Build a Web App with an AI Backend](https://learn.ynteractive.com/content/lovable-n8n-web-app-ai-backend) lesson, where one agent uses two tools: one to save a lead and one to email a reply. Last updated November 10, 2026 [Take the free module →](https://learn.ynteractive.com/courses/intro-to-ai-automation/8)Module 8 of the free [Intro to AI Automation](https://learn.ynteractive.com/courses/intro-to-ai-automation) course, about 15 minutes. No account needed. By [Robert Breen](https://learn.ynteractive.com/content/about) · Published November 10, 2026 · Updated November 10, 2026 · 6 min read n8n · AI Agents ## Watch: the idea in 0:47 ### n8n AI agent tools Explainer · 0:47 · with captions n8n AI agent tools explained: Google Sheets Tool and Gmail Tool vs regular nodes, naming tools in the system message, and limiting a tool to drafts. **Transcript** ChatGPT answers you in text. An n8n agent can also act, with tools. A chatbot just talks. An agent does things. A regular node runs every time, in order. A tool runs only when the agent decides to use it. That's the difference between a fixed workflow and an agent. Tell the agent when to use each tool. The lesson's system message is three lines. Add the lead with the Google Sheets tool. Reply with the Gmail tool. Each tool gets its own credential. Then every form submission is saved as a row, and gets a reply email. Give each tool the smallest job that works. The email agent's Gmail tool opens set to send. Change it to Draft. Then the worst it can do is write a draft you never send. Try it yourself, free, in Lovable + n8n. No code needed. ## What is a tool in an n8n AI agent? A tool is an action the agent can take, plugged into the **Tool** connection under the [AI Agent node](https://learn.ynteractive.com/content/glossary/ai-agent-node). The lessons put it in one line: "A chatbot just talks. An agent does things." Without tools, an agent can only answer in text. With a Google Sheets Tool, it can write rows; with a Gmail Tool, it can create drafts or send email. An AI Agent in the lessons is built from four parts: - **Chat Model**, the brain (an OpenAI Chat Model such as gpt-5-mini or gpt-4o). - **Memory**, such as Simple Memory, so it remembers the conversation. - **Tool**, one or more actions it can take. - Its instructions, in the **System Message** option on the agent itself. ## How is a tool different from a regular n8n node? A regular node always runs, in the order you connected it; a tool runs only when the agent decides to call it. That's the core difference between a fixed workflow and an agent. | | Regular node | Tool under an AI Agent | | --- | --- | --- | | Example from the lessons | **Google Sheets: Get row(s) in sheet** and **Gmail: Send a message** in Daily Job Alerts | **Google Sheets Tool** and **Gmail Tool** in the agent lessons | | When it runs | Every run, in order | When the agent's instructions call for it | | Who fills in the values | You, with expressions | You, or the model with the **✦** button | | Good for | The same steps every time | Steps that depend on what the input says | The Daily Job Alerts workflow uses no AI at all: it reads the sheet, filters it and emails each tech, every morning, the same way. [AI Agent vs Chatbot vs Automation](https://learn.ynteractive.com/content/blog/ai-agent-vs-chatbot-vs-automation) explains the bigger picture. ## How do you add a tool to an n8n AI agent? Click the **+** under **Tool** on the AI Agent and pick the tool, then set its credential and operation. The [AI Receipt Extractor](https://learn.ynteractive.com/content/ai-receipt-extractor-n8n-google-sheets) lesson's steps for the Google Sheets Tool: 1. Click the **+** under **Tool** and choose **Google Sheets Tool**. 2. Open **Credential**, choose **Create new credential**, leave it on **Managed OAuth2** and click **Sign in with Google**. 3. The tool opens set to **Get row(s)**, which reads a sheet. Open **Operation** and choose **Append Row**. The node on the canvas renames itself to match. 4. Pick the **Document** (the **Invoices** sheet) and the **Sheet** (**Sheet1**). 5. Click the **✦** next to each column so the model fills it in. ## How does the agent know when to use a tool? You tell it in the system message, naming the tool and what to do with it. The Lovable lesson's whole system message is three lines: ```text You are a helpful assistant. Use the google sheets tool to add the lead to the google sheet. Use the gmail tool to reply to the form submitter. ``` The receipt lesson does the same: "Use the google sheets tool to add the receipt to the sheet." The chat-based agents add a condition instead, such as "When the user asks you to save the ideas, use the Google Sheets tool to store each idea." That's why nothing is saved until you type "Yes, save those ideas to the sheet." Naming the tool matters because the lesson's voice-over points out that the system message alone does nothing: the agent "cannot do either until we give it the actual tools." Instructions and tools go together. ## Can an n8n agent use two tools at once? Yes. Add a second tool under the same agent and mention both in the system message. In the Lovable lesson, after the Google Sheets Tool, you click the **+** next to it and add a **Gmail Tool**, so every form submission is saved as a row and gets a reply email. ## Try it free, step by step **Lovable + n8n: Build a Web App with an AI Backend**. Build a contact form in Lovable and connect it to an n8n workflow — webhook trigger, AI Agent, Google Sheets, Gmail, and a Respond node that ties it all together. You do every step yourself in a practice copy of Lovable and n8n, it checks your work as you go, and “Do it for me” finishes any step you get stuck on. Free, and nothing touches your real accounts. [Start the free lesson](https://learn.ynteractive.com/modules/lovable-n8n) [Or read the full step-by-step guide](https://learn.ynteractive.com/content/lovable-n8n-web-app-ai-backend) Each tool needs its own credential. In the lesson, Gmail gets its own Google sign-in and a longer list of permissions than Sheets. When the workflow runs, the agent calls both tools and then the **Respond to Webhook** node tells the form it worked. ## How do you stop an agent's tool from doing too much? Set the tool to the smallest action that does the job. In the [AI Email Responder](https://learn.ynteractive.com/content/ai-email-responder-gmail-drafts-n8n) lesson, the **Gmail Tool** opens with **Resource** set to Message and the operation set to send. The lesson changes it: 1. Open **Resource** and choose **Draft**. The **Operation** switches to **Create**. 2. Set **Subject** to `Re: {{ $json.Subject }}`. 3. Click the **✦** next to **Message**, so the model writes only the body. 4. Under **Add option**, set **Thread ID** to `{{ $json.threadId }}` and **To Email** to `{{ $json.From }}`. As the lesson puts it, "the worst this agent can do is write a draft you never send." The system message repeats the limit in words: "You cannot send email. You can only create drafts." [AI Email Replies Without Auto-Sending](https://learn.ynteractive.com/content/blog/ai-email-replies-gmail-drafts) covers that build in full. The same thinking applies to Sheets: an agent that only needs to add rows should use **Append Row**, not an operation that changes or deletes existing ones. ## When should you use a regular node instead of a tool? Use a regular node when the step should happen the same way every time and needs no judgment. Sending each tech their jobs at 7am is a fixed sequence, so the job alerts lesson uses plain Google Sheets and Gmail nodes and no model at all. It's cheaper, faster and easier to test. Use a tool when what happens depends on the input: what a customer asked, what a receipt shows, which of several actions makes sense. A good rule: if you can write the steps as a checklist with no "it depends," build a fixed workflow. On cost: the lessons are free and run in practice copies. A real agent needs n8n (n8n Cloud is paid after a free trial, or self-host it), a Google account and an OpenAI API key, which is pay-as-you-go and billed separately from ChatGPT plans. [n8n AI Agent Tutorial](https://learn.ynteractive.com/content/blog/n8n-ai-agent-tutorial) builds a first agent step by step, and the [n8n tool page](https://learn.ynteractive.com/content/tool/n8n) lists every n8n lesson. For examples of finished client projects, browse the [n8n case studies](https://ynteractive.com/case-studies), or [hire an n8n consultant](https://ynteractive.com/hire-an-n8n-developer) to build an agent with your own tools. ## Key takeaways - An n8n tool is a node attached under an AI Agent's Tool slot, and the agent decides when to call it. - A regular n8n node runs every time in a fixed order, which suits steps that need no judgment, like daily job emails. - The system message names each tool and when to use it, such as "Use the google sheets tool to add the lead to the google sheet." - One agent can use several tools, such as a Google Sheets Tool to save a lead and a Gmail Tool to reply. - Setting the Gmail Tool to Draft and Create instead of send limits the agent to writing drafts a person reviews. ## Frequently asked questions **What tools can an n8n AI agent use?** The lessons use the Google Sheets Tool and the Gmail Tool. n8n offers many more; you add each one under the agent's Tool connection. **Does the agent use a tool every time it runs?** Only when its instructions call for it. A chat agent told to save "when the user asks" won't write rows until you ask. **Can I stop an n8n agent from sending emails?** Yes. In the Gmail Tool, set Resource to Draft so the operation is Create, and say in the system message that the agent can only create drafts. **Do I need memory to use tools?** Not always. A chat agent needs Simple Memory so "save those" refers to earlier messages; a webhook agent that handles one form at a time doesn't depend on it. ## Free lessons in this post - [Lovable + n8n: Build a Web App with an AI Backend](https://learn.ynteractive.com/content/lovable-n8n-web-app-ai-backend): Build a contact form in Lovable and connect it to an n8n workflow — webhook trigger, AI Agent, Google Sheets, Gmail, and a Respond node that ties it all together. - [n8n AI Email Responder: Draft Gmail Replies Automatically](https://learn.ynteractive.com/content/ai-email-responder-gmail-drafts-n8n): Build an n8n agent that watches your Gmail, reads each new customer email, and writes a reply as a draft in the same thread, ready for you to check and send. - [AI Receipt Extractor: Receipts to Google Sheets with n8n](https://learn.ynteractive.com/content/ai-receipt-extractor-n8n-google-sheets): Drop a photo of a receipt into an n8n chat. An AI agent reads it, pulls out the date, vendor and amount, picks a category, and adds a row to Google Sheets. ## Watch it in a free lesson: AI Receipt Extractor: Receipts to Google Sheets with n8n ### AI Receipt Extractor: video walkthrough Full walkthrough · 5:35 · with captions Drop a photo of a receipt into an n8n chat. An AI agent reads it, pulls out the date, vendor and amount, picks a category, and adds a row to Google Sheets. The whole lesson narrated step by step, recorded in Stepthrough’s simulated practice environment. [Watch it with chapters and the full transcript](https://learn.ynteractive.com/content/ai-receipt-extractor-n8n-google-sheets#watch) From the free lesson [AI Receipt Extractor: Receipts to Google Sheets with n8n](https://learn.ynteractive.com/content/ai-receipt-extractor-n8n-google-sheets) ## Related terms Plain-English definitions, each with an example from a free lesson. [AI agent](https://learn.ynteractive.com/content/glossary/ai-agent) · [AI Agent node](https://learn.ynteractive.com/content/glossary/ai-agent-node) · Chatbot · Accounts payable · [Action](https://learn.ynteractive.com/content/glossary/action) · [Agent memory](https://learn.ynteractive.com/content/glossary/agent-memory) · [AR aging report](https://learn.ynteractive.com/content/glossary/ar-aging-report) · [Artificial intelligence (AI)](https://learn.ynteractive.com/content/glossary/artificial-intelligence) [The full AI glossary](https://learn.ynteractive.com/content/glossary) ## Related prompts - [AI agent system message prompt](https://learn.ynteractive.com/content/prompts/agent-system-message-draft-prompt): An AI agent prompt template for the n8n System Message: who the agent is, what it knows, its rules and how it writes, plus a… - [Daily priorities plan prompt](https://learn.ynteractive.com/content/prompts/daily-priorities-plan-prompt): Paste your to-do list and fixed times and get a realistic plan for today: top three, time blocks, a buffer, and what to move to… - ["Do you service my area?" reply prompt](https://learn.ynteractive.com/content/prompts/service-area-question-reply-prompt): Answer "do you cover my town?" from your real service area list: a clear yes with how to book, a kind no, or a request for the… - [FAQ from customer emails prompt](https://learn.ynteractive.com/content/prompts/faq-from-customer-emails-prompt): Turn a batch of customer emails into an FAQ: the questions people really ask, in their words, answered only from your facts… [All free prompt templates](https://learn.ynteractive.com/content/prompts) ## Visual guides The same ideas as quick visual guides: swipe through the slides, or read each one as text. - [Build a web form with an AI backend, no code](https://learn.ynteractive.com/content/visual/lovable-n8n-form): Most website forms just land in an inbox and wait. This one does the work. You describe the form in Lovable, and it builds it. - [Receipts to Google Sheets with an AI agent in n8n](https://learn.ynteractive.com/content/visual/receipt-extractor-spot-check): A box of paper receipts at the end of the month. Typing them in is slow and easy to get wrong. - [Automate Email Responses with AI: Drafts-Only System Message](https://learn.ynteractive.com/content/visual/email-agent-system-message): Automate email responses with generative AI without auto-sending: the system message from the AI Email Responder lesson that answers FAQs as drafts. [All visual guides](https://learn.ynteractive.com/content/visual) ## Keep reading - [Learn n8n for Beginners: 6 Free Projects in Order (2026)](https://learn.ynteractive.com/content/blog/learn-n8n-for-beginners): Learn n8n for beginners with six free, hands-on projects in a sensible order, from templates to AI agents to a web form, plus what you need for real. - [n8n Defined Automatically by the Model, Explained](https://learn.ynteractive.com/content/blog/n8n-defined-automatically-by-the-model): What "defined automatically by the model" means in n8n tools, when to click the sparkle button, and when a fixed expression is the safer choice. - [n8n Simple Memory Explained for Beginners](https://learn.ynteractive.com/content/blog/n8n-simple-memory): n8n Simple Memory explained: what it keeps, why "save those ideas to the sheet" needs it, how email agents key it by thread, and when an agent needs none. [All blog posts](https://learn.ynteractive.com/content/blog) · [All free tutorials](https://learn.ynteractive.com/content)