Our Dendrite CLI enables you to securely to share your website permission with your AI agent, mirroring the access you have in your local browser. Everything is stored locally ensuring that no one else other than you has access to the authenticated sessions.

1

Run Command

First, run the following command in the terminal dendrite auth --url <url> where url is the website you want your agent to be authenticated against. For example dendrite auth --url mail.google.com. This will open up a browser at that url.

2

Login and Save Auth Session

In the newly opened browser login into the service. When you are done go back to the terminal where you ran the command and press enter to save the seession. extension icon and press Save authentication session

3

Authenticate in Script

Once your session has been saved, you can use it in your Dendrite browser like this:

Dendrite(auth="domain.com")
Dendrite(auth=["domain1.com", "domain2.com"])

​
Example: Authenticating on Instagram

To log in to instagram with your agent your need to:

1

Log in to Instagram in your Local Browser

2

Save your Authentication Session

Use the Dendrite CLI to save an authenticated session with Dendrite. Run the following command dendrite auth --url instagram. and login with your credentials.

3

Use with Dendrite

Initiate your Dendrite browser with an authenticated session.

auth_example.py
# Initiate browser with instagram session
browser = Dendrite(auth="instagram.com")

# Go to website with authenticated session
browser.goto(
  "https://instagram.com",
  expected_page="Logged in version of Instagram feed"
)
auth_example.py
from dendrite import Dendrite

browser = Dendrite(auth="instagram.com")

browser.goto(
    "https://instagram.com",
    expected_page="You should be logged in with the instagram feed visible",
)
desc = browser.ask("Describe the first post in my feed")
print("First post description: ", desc)

​
Let’s scale up

Remote Browser

Use a remote browser to run web automations in the cloud