Run stock screens with the crible CLI

Inside the repo, run every command with uv run crible … (run uv sync once if the environment is missing). Outside the repo, crible is an installable tool — uv auto-provisions the required Python 3.12, and bootstrap pulls the FULL published dataset (zero crawl), so three commands stand up a working screener anywhere:

uv tool install git+https://github.com/maxgfr/crible
crible --data-dir ~/.crible-data bootstrap
crible --data-dir ~/.crible-data screen "piotroski_f >= 7" --sort -composite_rank

(One-shot alternative: uvx --from git+https://github.com/maxgfr/crible crible ….)