Your biggest issue here is that your config file is full of cut and pasted directives that would make it impossible to debug. e.g. why are you disabling sendfile? You seem to have lots of config for bypassing a local cache, but no local cache defined. Please learn how to use the include directive to make this more manageable and get rid of the redundant stuff.
I tried to add some headers but they are completely ignored
You didn't tell us what you tried. IME ngx_http_headers_module comes as standard in most distributions of nginx and add_header works exactly like its supposed to.
Another issue is related to /wp-admin folder which I can't access. It enters in some loop with response code 302.
Probably redirecting to the URL you just requested? (it would have been helpful if you had provided the URL you were requesting and respose headers). That's because wp_admin checks to see if the incoming connection is over https. Yours is not. Your SSL appears to be terminated on a reverse proxy (again, my crystal ball is working overtime). A quick Google turned up https://wordpress.stackexchange.com/questions/387990/how-do-i-handle-ssl-properly-when-wp-is-behind-a-reverse-proxy and http://www.michaelfoster82.co.uk/how-to-set-up-wordpress-behind-a-secure-reverse-proxy-using-nginx/
I don't know why but some resources are loaded via http and some others via https
Sloppy programming. But making wordpress believe that the incoming connection is over HTTPS (see above) should fix.