Browser Logins
Use browser automation to connect to services without APIs
Browser Logins
Ok, now it gets interesting. If the service you want to retrieve data from or write data to doesn’t expose an API, you can use a browser connection. When you create a browser connection, Pinkfish will open a browser window to the login page of your selected service. You will then authenticate, solve any challenges (like captchas, one time passwords, or 2fa), and then Pinkfish will store your logged in session details. Note that Pinkfish does not store your login credentials, only your session details.
Once saved, you can use that saved session in tandem with Pinkfish’s browser automation skills. Your saved browser Connections will be available in your slash command menu list (just like your API Connections).
The two skills that support browser Connections are /scraper and /browser session.
Scraper
Scraper allows you to navigate to a given page and to scrape the contents. It supports LLM Vision scraping, so you can get really specific about what data you want to bring back and in what format.
Browser Session
Browser session is a more full featured browser automation skill that allows you to do things like fill out forms, click buttons, download files, and more.
Example prompt
Here’s an example showing use of a browser connection to scrape a web page.
In this example, I selected my “Babylon” browser connection and then used the /scraper skill to scrape my private profile page, which I can only access as a logged in user.
Session expiry & Re-authentication
Browser sessions work the same as if you’d logged in to the service yourself. They will expire whenever the service’s session expires. Sometimes it’s a few hours, sometimes its a month. If the service offers a “remember me” option, you can use that to extend the session timeout. You’ll need to monitor your browser workflows and re-authenticate as needed.
To re-authenticate, open the browser connection and click the “Update Connection” button. Re-login and you’ll be good to go.