install.python-poetry.org returns 403 (Cloudflare error 1010) to Python's urllib since the move to Cloudflare Pages #192

Since the hosting move to Cloudflare Pages on 2026-11-08 (#189), https://install.python-poetry.org answers 403 Forbidden with body error code: 1010 when the request uses Python's default urllib User-Agent (Python-urllib/3.x). Other clients still get 200.

This breaks the python -c 'from urllib.request import urlopen; print(urlopen("https://install.python-poetry.org").read().decode())' | python - pattern that many Dockerfiles use. The download fails and python - receives an empty script, so the build only fails later with poetry: not found (exit 127), which makes the cause hard to spot.

Reproduce (2026-11-10 15:07 UTC, Python 3.12.14):

$ python3 -c 'from urllib.request import urlopen; urlopen("https://install.python-poetry.org")'
urllib.error.HTTPError: HTTP Error 403: Forbidden

$ curl -s -o /dev/null -w '%{http_code}\n' -A 'Python-urllib/3.12' https://install.python-poetry.org
403
$ curl -s -A 'Python-urllib/3.12' https://install.python-poetry.org
error code: 1010

$ curl -s -o /dev/null -w '%{http_code}\n' https://install.python-poetry.org          # curl's default UA
200
$ curl -s -o /dev/null -w '%{http_code}\n' -A 'python-requests/2.32' https://install.python-poetry.org
200

One blocked request's cf-ray, in case it helps: a585f78a0c2e8923-YHZ.

Error 1010 is Cloudflare blocking the client's browser signature. A zone-level setting (Browser Integrity Check, or a bot/WAF rule) seems to be catching Python-urllib, probably as a side effect of the move rather than on purpose. Could it be turned off or relaxed for this host? Until then, sending any other User-Agent works around it, e.g. urlopen(Request(url, headers={"User-Agent": "curl/8"})).

Thanks for maintaining the installer!