A founder was sharing her screen during a work call with a CEO. He asked her to check a lead's website. She searched the domain, clicked it, and an adult website opened on the shared screen.
That specific incident came from a public post on X, and it is a much better product brief than most I have written for myself. Here is the post that prompted this build.
It is easy to laugh at a one-off mishap. It is harder to dismiss the pattern. A domain can be mistyped, a search result can be misleading, a link can redirect, or the right page can expose the wrong thing at exactly the wrong time. A shared screen removes the usual private moment between clicking and seeing the result.
For people who demo products, support customers, teach, interview, or work with clients, that is a small but real risk surface. It is not only about inappropriate websites. It is also a personal inbox, a pasted token, a payment detail, an API key, or an unfamiliar domain that imitates a brand you trust.
So I built ScreenGuard, an early Chrome extension that puts a small safety layer between a click and an embarrassing moment.
What it does
ScreenGuard has two related jobs.
First, it warns before a risky destination opens. It looks for signals such as known unsafe or adult domains, phishing-style lookalike domains, and suspicious navigation patterns. The point is not to declare every unfamiliar link dangerous. It is to create a brief, deliberate pause when a link deserves one.
Second, it has a screen-share protection mode. When enabled, it can blur sensitive values on the page, including things such as API keys, tokens, email addresses, and payment-card-style numbers. A small indicator tells you how many items are hidden, and you can reveal them when you need to.
The privacy rule matters here: the extension runs locally in the browser. ScreenGuard does not send browsing history or page contents to a server to make these decisions.
Why I did not make it a "block everything" tool
False positives are the fastest way to make a browser extension useless. A warning that fires on every unfamiliar domain gets trained out of a user's attention in days.
That is why the design is layered. Known dangerous destinations can be blocked under the relevant settings. Heuristic matches get a warning and a choice. Sensitive-content protection is specifically tied to screen-sharing mode, rather than permanently hiding useful information during normal work.
The product has to make risky moments less likely without turning everyday browsing into a sequence of permission dialogs.
What I am looking for now
ScreenGuard is free and currently at v1.0.15. I am looking for early feedback before I broaden the feature set.
I would value comments from people who regularly share their screens:
- What is the worst thing you have accidentally shown during a call or demo?
- Which protection would earn a permanent place in your browser: suspicious-link warnings, adult-site blocking, phishing checks, or sensitive-data blurring?
- Where would a warning be genuinely useful, and where would it become annoying?
I am especially interested in feedback from developers, founders, sales and support teams, recruiters, educators, and freelancers. If the problem is real but the current solution misses the workflow, I would rather learn that now than build a larger product around the wrong assumption.
What would you want a screen-sharing safety layer to catch before it catches you?
