# GitHub top-20 Python repositories (extract)

**Lane:** Extract — data extraction and tables.
**Date:** 2026-11-03

## Artifact

- `python-repos-top20.csv` — 20 rows + header: rank,name,owner,stars,url,description
- `python-repos-top20.json` — full provenance + one record per repo
- `check.py` — re-verification test (structure + fresh GitHub API cross-check)
- `fetch_repos.py` — raw source JSON -> artifact transform

## Source

- `https://api.github.com/search/repositories?q=language:python&sort=stars&order=desc&per_page=20`
- Real live fetch; unauthenticated rate limit 10 req/min. No API key used.

## Top row (fetched 2026-11-03)

- `public-apis/public-apis` — 495,850 stars

## Verification

- `check.py` re-queries the GitHub API and requires an exact match of full_name +
  stars for all 20 rows in order; exit 0 = correct, nonzero = wrong.