Build N8n Workflows With Firecrawl Interact ![rw-book-cover](https://img.youtube.com/vi/jAbhIRC371s/maxresdefault.jpg) Hey there, it's Leo from Firecrawl, and in this week we'll be taking a look at how you can build an entire N and workflow using the latest features from Firecrawl. Instead of explaining each thing that we'll be creating, let's just dive straight to it and actually start building the workflow. So, for the first example, let's add a manual trigger. This will execute every time I click this, and now let's add the Firecrawl scrape. You always have to scrape before interacting, since to interact with the page is basically a fallback of that initial scrape. Here is where you'd place the URL. You can change that to whatever you'd like, but with firecrawl.dev, let's execute that. Already done. This is the result I got back. If you scroll all the way down, you should find the scrape ID listed down here. Let's pin this, and now create another node. Firecrawl, then let's scroll down and select execute interaction. Then just find the scrape ID from the past execution, drag it over there so that this node understands which ID it should be using. And now inside of the interact operation, I can either choose to execute code or prompt. Since I don't have this attached to an actual agent, prompting is what we're going for, because this will use an agent to go inside of the website and fetch what we need. So, if I add something here like list the pricing plans. So, pretty simple, let's execute this step and wait for it to go inside of the website, scroll down to the pricing section, which we can actually follow along here in the dashboard. On the right tab, you can find the interact live view URL. Open that up and see, as well as interact with the current browser session that your agent has access to. But since it already found the pricing page, we might have that data already done here. So, now in the output parameter, you should see the response. And by the way, one of the parameters we get as a response is the actual live interaction view URL that we just accessed back here. And what we mean about it being interactive, if you haven't seen our past videos, is that you can actually go inside of this browser, click to sign up, click to log in, and authenticate yourself inside of the session so that your agent has access to the same data you're seeing behind the authentication. Now, for this simple workflow, the only thing missing is to actually close that session after you've gathered the data. So, just go down to interact actions, select stop interaction, and then grab the scrape ID here as well. So, just find that and drag it over. All right, now that I execute this step, I can go back to sessions and see it being deleted. Okay, node executed successfully, and 3 2 1, it's gone. All right, this is just one way you can use the interact feature, mainly through the prompting, but you can also design your own agent to act inside of the interact and follow up with its own code inside of the website you need to scrape. So, let's build that right now. First, let me start by adding the AI agent node. And this node needs a chat trigger. So, let's find that. Let's get this trigger down here and make the initial connection. Let's add a chat model. I like using OpenRouter, so let's choose that, as well as Claude Sonnet 4.6. Now, for the tool, let's first add scrape from Firecrawl. Again, we'll need to tell the agent that it first needs to execute scrape to get that ID and then use the ID inside of the interact feature. Firecrawl, Firecrawl tool, and from these tools, let's select scrape. It's already selected by default. The only thing we'll have to do is make sure the LM selects these specific fields. All right, now with that done, let's select the interact tool. Go over to resource, find interact, and the first thing we'll want to add is the interact context. If you're familiar with skills, this is basically a huge instruction for the system prompt, so that it only triggers this, only uses this specific tool when it needs to use the interact feature. Most of the times you want to scrape a website, you won't need to actually interact with it. So, just having scrape inside of the context window won't waste so much tokens. But if you need to interact, you should provide enough context for the LM to operate with the set of codes inside of those pages to then find each specific components to then interact with. And so, because we have this tool, what we need inside of the system prompt is something as simple as telling it we have the scrape tool, and then just following up with if the website has heavy JavaScript, then consider using interact. But finally, a single line that describes always use the interact context tool to get full context of how to use the interact tools. With that done, let's actually add the interact tool. So, go over to resource, interact, now execute interaction. Let's make this dynamic, as well as the opt dynamic. Here is where it'll be different from the previous workflow, because I don't want an agent telling another agent what it needs, because this makes it a bit more complex and unnecessary. So, what we want to add here is code. It will rather be directly executing code inside of that browser session. So, let me click to let the model define this parameter, and your execute interaction node should pretty much look like this. Let me drag this down here and now add the tool to stop the interact session. So, resource, interact, operation, stop interact, and then the LM should be able to add the scrape ID of the session that should be stopped. Okay, with this done, we can do our classical example, which is to have the LM go over to this website, click the see more button, go to the second page, and basically grab all the 21st rows from this table. So, here's my prompt. Go to airank.dev and get me the top 20 results from the homepage table. Let me hit enter, and now it first tried to scrape the page, understood it needs to use interact, and then the fallback was actually having the interaction. So, if we go back to sessions inside of the Firecrawl dashboard, we should see that session running inside of airank.dev. There it is. Let's open the live view to see exactly what it's doing right now. All right, apparently it already clicked on the see more button, and now it found the second page button, already understood it grabbed the data, is retrieving all that as context to the AI agent, which should be done in just a few seconds. All right, it's done, and here are the results. By the way, this was the first execution for that AI agent, and it took around 1 minute and 41 seconds, just because it actually needs to perform actions inside of a webpage to fetch for that specific data. But I believe this really empowers enrichment workflows to get past specific operations that only can be handled through actual interactions. Now, the best part is that you can use all this for free in up to 100,000 credits. If you were to buy this inside of Firecrawl, it would be $83 a month, but you're getting all this for free inside of N and with the quick connect feature. To access this, just go over to credentials for Firecrawl. Up here, you should see a setup manually. Just toggle this over to use quick connect. After selecting that, just click connect to Firecrawl, and you should get the free hobby plan along with 100,000 free credits. And if you're wondering how many credits you'll be spending, for each scrape, it costs just one credit. Then following up with the interact, it costed 27 credits. And then for the example using your own agent, since that cost with the LM is all yours, the credit expense for the interact was down to three credits. If you have used our agent feature before, you know this is a considerably lower amount of tokens you need to spend. So, yeah, let us know which workflows you're building. Thanks for watching, and we'll see you in the next video. Till then., Build N8n Workflows With Firecrawl Interact, YouTube, 01kn56kqbtvx392h3ey4zwprc2.