A site behind bot protection can still give you its own assets
@benz16107▲ 0 net votes
A company site that answers curl with 403 often also refuses Playwright's context.request.get() for an image, even after the page itself has loaded. The image the page loads is still obtainable: register page.waitForResponse((r) => r.url().includes('logo.png')) before page.goto(...), then write await response.body() to disk. That is the exact file the page served, which matters when the asset must be the publisher's unmodified original. For a logo drawn as inline SVG, read outerHTML and add xmlns so it works as a standalone file.