A captcha/login page which only allows you to register if you post an image of you touching grass
  • Jinja 62.4%
  • Python 37.4%
  • CSS 0.2%
Find a file
2026-05-16 22:35:22 +02:00
static First commit 2025-08-08 18:08:41 +02:00
templates remove english words and use numbers instead, fix XSS 2025-10-24 18:08:15 +02:00
.gitignore fix the word embedding cache being included in git and try switching to v3 which might fix jina being so slow 2025-08-21 19:23:13 +02:00
.python-version First commit 2025-08-08 18:08:41 +02:00
constants.py remove english words and use numbers instead, fix XSS 2025-10-24 18:08:15 +02:00
CREDITS First commit 2025-08-08 18:08:41 +02:00
jina.py try using jina clip v2(v3 is not multimodal) 2025-08-21 19:25:31 +02:00
LICENSE First commit 2025-08-08 18:08:41 +02:00
main.py remove english words and use numbers instead, fix XSS 2025-10-24 18:08:15 +02:00
ocr_check.py remove english words and use numbers instead, fix XSS 2025-10-24 18:08:15 +02:00
pyproject.toml Update libraries to latest versions 2026-05-16 22:35:22 +02:00
README.md fix register not working due to wrong url_for parameter and update README.md for a better solution to easyocr 2025-08-12 15:51:15 +02:00
requirements.txt Update libraries to latest versions 2026-05-16 22:35:22 +02:00
uv.lock Update libraries to latest versions 2026-05-16 22:35:22 +02:00

A captcha/login page which only allows you to register if you post an image of you touching grass

To install, you have to install everything in requirements.txt or do uv sync, but you also have to install easyocr. On my machine, i did this because i wanted to install the CPU version of torch or otherwise it would install NVIDIA bloat. If you want to do the same, add extra-index-url:

uv pip install easyocr --extra-index-url https://download.pytorch.org/whl/cpu

Dont add the flag if you have a dedicated GPU on the computer where you host the project. Run the command without uv if you use pip.