Vanta Agent now grabs your compliance screenshots for you
One of the most tedious parts of compliance just got a whole lot easier.
Tired of hunting down a screenshot, then manually uploading it to the right evidence request? Yeah, us too.
Now, just ask the Agent instead; something like "can you find my company's privacy policy from [website] and upload the necessary evidence?" and it takes care of the rest: reads your open evidence requests, finds the right pages, captures the screenshots, and shows them to you before anything goes anywhere. Nothing gets submitted without your approval. Once you give the thumbs up, it lands in the right evidence request automatically. No tab-switching, no manual uploads. Just say what you need and confirm it looks good.
This is rolling out over the next couple weeks, but if you'd rather see this live sooner, react with a raised hand, and I'll get your interest submitted! π
hey Steve F. Allyson K. Rob G. Mark O.: this has been turned on for you! open up vanta agent and try it out. please drop any feedback you might have here in thread. Andrew K. Kasey C. Glenn G. Alyssa L. Esmeralda O. Jonathan M. i see your interest! standby... love seeing all this interest (i dont blame you! 10/10 feature for sure)
Using this now.
It's an amazing feature, can definitely see this being useful (especially running as a custom agent in the future!
)
Feedback:
It looks like the screenshots are cropped - would be great if this was a full page screenshot.
The agent wasn't able to get a screenshot of my cookie consent banner, it kept refreshing the page when the banner didn't show.
For #2, I tried taking control and opening the cookie consent banner directly and resuming the agent (the agent closed the prompt and went back to refreshing the site), which didn't work. I stopped the agent and steered it towards the cookie consent link in our website footer, which then worked.
Also I'm curious how/if this will work when we need to get screenshots for Github Changes/Terraform Code/ Jira tickets- how does it work with systems that need to be authenticated into? (like Mark said it would be dooope if we could build an agent to get sample screenshots and automatically upload them as audit evidence)
hey Mark O.:
full page screenshot: This is the team's next action item, it's coming soon!
2. screenshot of cookie consent: theyre looking into it and have logged it as feedback.
hey Allyson K.:
timezone: this is server time, so EST (AMER), but the team is looking into adding the timezone part to the time to make it clear github q, here's what the team shared: The agent will stop at the login page and prompt the user to take control and enter their credentials. Once they are logged in, they can resume the agent.
oof! lemme investigate π
This sounds super interesting. Are there plans to have this work with internal (non internet facing) tools?
Johan Andrew K. Esmeralda O. Patrick W. Kasey C. Jonathan M. activated!
Allyson K. still investigating the time thing.
and Nick D. what tools specifically? (PM is asking). Or are you saying generally non-internet spaces?
In general. For example it would be nice to pull in screenshots from internal wiki articles, internal tools or even external tools behind auth. In cases where integrations donβt offer data for custom tests it would be helpful.
