Add audit goal?

[thejcannon]

Is your feature request related to a problem? Please describe.
Goals like check and lint are code-related IMO, but we want to integrate tools and workflows that aren't code-related into automated CI (e.g. #13770).

Describe the solution you'd like
We should house these in a new goal, pluggable like the others.

Describe alternatives you've considered
N/A

Additional context
I decided to write this issue because I want to find dead 3rdparty top-level deps for pruning. ./pants audit seems like a perfect candidate for this.
Alternative names welcome.

[stuhood]

Given that this needs to always hit the network, and thus is inherently flaky (unlike lint and check), I think that I agree.

[thejcannon]

@stuhood specifically you mean pip-audit, right? I can think of auditing plugins that aren't network-dependent (" I want to find dead 3rdparty top-level deps for pruning")

[stuhood]

@stuhood specifically you mean pip-audit, right?

Yea. And cargo audit, etc.

I can think of auditing plugins that aren't network-dependent (" I want to find dead 3rdparty top-level deps for pruning")

Hm, interesting. It seems like as with lint and check, a potentially relevant bit is which targets might impact the goal... in that case, targets without dependees (or more generally "all targets") need to run, so the --changed flags aren't relevant (unless we came up with a way for --changed to detect targets which lost deps... quite hard to do though).

So that's probably at least two differences... whether those two differences ("intentionally-unstable-network-consuming", "always global") should represent one new goal or two is interesting.

[Eric-Arellano]

I continue to not love the name check even though I agree with our recently-agreed-upon definition that it's either type checking or the lightest form of compilation. Here, I think the difference between audit vs check vs lint is not super obvious. But, I do agree audit-type checks are better kept separate from lint/check, and I think audit is a solid name.