Sign up. Free, no card. Then ask Claude, Cursor, or ChatGPT to run the loop on this page.

PyPI stats agent

No PyPI board. Start with pandas. Run source python 3M and 12M. Repeat for polars. If the agent sends source pypi, that step errors and must be retried as python. sklearn is wrong when the project is scikit-learn.

The source string is python. That is the whole trap.

Project name from PyPI, not the import.

Answer in install interest on PyPI, not GitHub stars.

The loop

Pull 12-month python growth for pandas and polars. Burst vs still installed.

1. get_growth(keyword="pandas", source="python", percent_growth=["3M","12M"])
2. get_growth(keyword="polars", source="python", percent_growth=["3M","12M"])
3. If source pypi errors, retry source python
4. Answer. Burst vs still installed.