Some startups are run by cowboys/cowgirls. They don't care about anything other than features until everything falls apart or they rebuild in a nightmare sprint.

I've been at one of these. You will fail to change them even if you stay for 3 years. You risk getting fired if the CEO realizes the mess and hires in someone to fix it. Your manager is already thinking about firing you.

Apply for new jobs immediately. In the meantime do 1 pr a day.

For next time: it's really easy to accidentally ruffle feathers if you come in and start proposing tooling or process changes. People get attached to their project if they've worked on it for a while, and can take observations from an outsider as judgemental or an attack, regardless of how valid those observations are (many of yours seem valid to me). You coming from big tech can make people even more sensitive to this (if you're triggering their latent impostor syndrome or whatever as someone from a prestigious company). The way you summarize your manager's feedback makes me think that you either rubbed him the wrong way or offended some more tenured people who complained. Kind of hard to undo that now, but helpful to keep in mind for the future. If you demonstrate that you can hack it on tickets as you ramp up and establish some trust with your teammates over a few months, you'll typically have an easier time with this type of stuff.

(I would normally add a caveat that another good reason to wait a bit before bringing stuff up is that you might be missing the bigger picture, and that can help you ensure that you're actually arguing for useful change. I omit it for you because "we should have any observability at all for our production application" seems pretty uncontroversial to me and doesn't make me think you're out of touch or focused on unimportant things)

Poltics goes a long way in this. You need to make others happy first and build up some goodwill first. If you want to tough it out here, do this first, make some allies, and then slowly suggest changes.

Don't suggest to everyone at once, try to build consensus for your changes one engineer at a time. Ideally you'd make friends with the lead or others you can pitch to first privately.

But, having been in a somewhat similar situation, I can tell you that some companies and teams are just dysfunctional and "cliquey". It has nothing to do with anything rational or logical or something you did or didn't do, it's just broken beyond repair. My guess would also be that they have had a number of people come and go probably due to the same reasons.

That manager also sounds like they shouldn't be managing at all. Besides that silly metric(did they also specify how many lines of code they want per day? ridiculous), it would've been more useful if they explain why your changes were turned down at least. There might be a good reason that you're not aware of. A sane manager would choose to guide you to make sure your valuable contributions benefit the business and align with any strategy as opposed to talking to you like a child and going "do it because I said so". Your teammates may have been having a bad week or feel threatened but I blame your manager. They should know better.

I recommend you plan your exit as nothing good will come out of you staying.