I had Rust listed on my site for about three weeks. During that time exactly one artifact in my GitHub history had anything to do with Rust: an unmodified cargo new scaffold. Default Cargo.toml. Default main.rs. Hello world, untouched.
I noticed it during a repo audit — the kind where you go through every claim on your own site and ask "if someone clicked this, what would they actually find." For most things the answer was fine. For Rust, the answer was a scaffold I'd generated once, probably to see what the tooling felt like, and never touched again.
Here's the thing about skills lists: they're read as claims, not aspirations, even when you don't mean them that way. "Rust" on a resume reads as "I can write Rust," not "I once ran a command that generates Rust." Nobody scrolling a GitHub profile is going to check. Most won't click through. The ones who do are exactly the ones whose judgment you should want to survive contact with.
That's the asymmetry that made the decision easy. Keeping "Rust" on the list bought me approximately nothing. But it cost something specific: the one person who clicked through and found a default scaffold now has a concrete, verifiable reason to discount everything else on the page. Not "maybe he's exaggerating somewhere" — "I checked one claim and it was false, so what's my prior on the rest."
Credential inflation isn't a modesty problem. It's an epistemics problem: a false claim, once caught, reduces the evidential value of every true claim next to it. A site with ten accurate claims and one inflated one isn't 90% trustworthy to a skeptical reader — it's a site where you now have to independently verify all ten.
So I removed it, along with Solidity, CUDA, and MATLAB. What's left is smaller. It's also the first thing on the site I'd bet real money on if someone tested it.
The rule I run now: nothing goes on the list unless there's a repo, a commit history, or a shipped thing behind it. If the honest version of a claim is "I tried the tutorial once," that's not a skill. That's a Tuesday.