Creative Commons Catalog vs GitHub ReadMe Stats

Creative Commons's search catalog was handed off to WordPress.org and relaunched as Openverse; the old `api.creativecommons.engineering` host now 301-redirects to the current `api.openverse.org`. A live unauthenticated GET against `api.openverse.org/v1/images/` returned a 200 JSON result set with `Access-Control-Allow-Origin: *`, confirming keyless anonymous search still works today (rate-limited to 200 requests/day per IP), with an optional `/v1/auth_tokens/register/` OAuth2 client-registration endpoint (confirmed live, returning a 400 validation error rather than 404) for higher limits. The catalog itself and every indexed work are free to use under their respective open licenses.

GitHub Readme Stats generates SVG stat/language/pin cards for GitHub profile READMEs; its handler explicitly sets `Content-Type: image/svg+xml` and, per its own README, the shared public demo at `github-readme-stats.vercel.app/api` is "best-effort and can be unreliable due to rate limits and traffic spikes" (confirmed today: it returned a `DEPLOYMENT_PAUSED` 503) — the project itself is very much alive (58k+ GitHub stars, commits within the last week) and explicitly recommends self-hosting or a GitHub Actions workflow instead of relying on the shared instance. Neither the Vercel serverless handlers nor Vercel's platform add a CORS header by default, correcting cors from "yes" to "no". Free and open-source, self-hostable.