FYI, my solution to my trust issues has been to accept them, and attack the other side of the trust coin: I remain as untrusting as before, I simply care less about the failure modes.

For me personally, this makes things a lot more entertaining. It becomes a game of seeing if I can predict if someone is going to violate my trust or not - I trust them with something trivial and unimportant to see what will happen. I expect and anticipate failure a lot of the time from my lab rats, and am usually pleasantly surprised when I don't get it.

Over time I build up models of the people I can trust in various ways, but even with the best of them it winds up looking like airport landing times: even the most reliable route fails fairly often, but as long as it doesn't cause cascading effects that's ok.