It’s not “assumed” to be secure.

It’s out there and visible for all to see. Hopefully, someone knowledgeable has taken it upon themselves to take a look at the software and assess its security.

The largest projects, like all the ones you named are popular enough that there’s no shortage of people taking a peek.

Of course, that doesn’t mean actual security audits are uncalled for. They’re necessary. And they’re being done. And with the code out there, any credible auditer will audit all the code, since it’s availiable.

Compare that to closed-source.

With closed-source, the code isn’t out there. Anyone can poke around, sure, but that’s like poking a black box with a stick. It’s not out there. You can infer some things, there are some source code leaks, but it isn’t all visible. This is also much less efficient and requires much more work for a fraction of the results.

The same goes with actual audits. Usually not all source code is given over to the auditers, so some voulnerabilities remain uninspected and dormant.

Sure, not having the code out there is “security”. If someone doesn’t see the code, it’s much harder to find the weakness. Harder, but not impossible.

There’s a lot of open-source software. There’s also a lot closed-source software, much more than the open-source kind, in fact.

What open-sourcing does is increase the number of eyes looking at the code. And each of those eyes could find a weakness. It might be a bad actor, but it’s most likely a good one.

With open source, any changes are publically visible, and any attempt to sneak a backdoor in has a much higher chance of being seen, again due to the large number of eyes which can see it.

Closed-source code also gives lazy programmers an easy way out of fixing or not introducing vulnerabilities - “no one will know”. With open source, again, there’s a lot of eyes on the code - not just the one programmer team making it and the other auditing it, as is often the case.

That’s why open source software is safer in general. Percisely because it’s availiable, attacking it might seem easier. But for every bad actor looking at the code, there’s at least ten people who aren’t. And if they spotted a voulnerability, they’d report it.

Security with open source is almost always proactive, while with closed source it’s hit-or-miss. Many voulnerabilities have to cause an issue before being fixed.