What the Agent can do

A look at what Buron's Agent can do during a conversation, from running reports to drafting documents.

When you talk to the Agent, it doesn't just respond with text. It can also reach out and do things, like pull a report from Google Ads, save a document to Buron, or search the web for something it doesn't know.

You don't have to think about which capability to use. The Agent picks based on what you ask. But it helps to know what's available.

Saving and reading docs

The Agent can read from and write to your team's docs in Buron, the same ones the CLI works with. Ask it to save a strategy doc, revise an existing one, or give you feedback on something you're drafting. The Agent handles the file work for you.

To work with these docs from your AI editor instead, check out Install the Buron CLI.

Pulling Google Ads data

When the conversation turns to your ad accounts, the Agent can list the accounts you've connected, pull data from any of them, and run the same health checks that power the Health tab. If you haven't connected an account yet, the Agent can walk you through that too.

These actions need an active Google Ads connection. To set one up, check out Connect your Google Ads account.

Looking things up on the web

For anything outside your accounts and your team's docs, the Agent can search the web and read individual pages. This is how it keeps up with what your competitors are doing or what's trending in your space.

Web results aren't fact-checked. Treat them like links a teammate sent you, useful as a starting point but worth verifying before you act on them.

Asking you for input

If the Agent needs more information from you, it'll pause and ask. You'll see the question right in the conversation, and you can reply inline. The Agent picks back up from there.

Approving an action

Anything that changes data outside of Buron needs your sign-off first. This includes pushing a change to Google Ads or updating an account setting. When that happens, you'll see an approval card with the proposed action. Click Approve to let it run, or Cancel to skip it.

Read the Agent's reasoning before you accept, especially the first time you see a new kind of action come up.

Seeing the work

As the Agent works, you'll see each step appear in the conversation. To take a closer look at what the Agent passed in or what came back, expand any step inline. This is the easiest way to understand why the Agent did what it did.