A CLI to take screenshots of websites and generate open graph images from your own computer. Having a website with a screenshot as an open graph image is a great way to promote your website and attract more visitors.
- Existing solutions are expensive and subscription based. You usually only need to take the screenshots once. You don't have to pay for a monthly subscription.
- Running your own open graph image generator in the cloud is also expensive. If you use a serverless platform, you'll have to pay for the computing power. If you use a self-hosted server, you'll have to pay for storage and bandwidth.
- Screenshot a single URL
- Capture screenshots of multiple URLs from a sitemap or RSS feed
Notes:
- Note that I only tested this on macOS.
- Chrome is required because I don't want to use puppeteer. Puppeteer automatically downloads a recent version of Chrome for Testing which can be ~170MB macOS, ~282MB Linux, or ~280MB Windows. That's a lot of bytes for doing just one thing. It turned out that the normal Chrome browser can be used for this task and majority of people already have it installed.
npm i -g og-screenshotsOr you can also directly execute the command:
npx og-screenshots --url "https://example.com" Usage
$ og-screenshots [input]
Options
--url, -u <type> Input URL (required). Can be a sitemap, RSS feed or a single URL. When using a sitemap or RSS feed, this command will automatically detect the URLs in the feed or sitemap and process them.
--transform, -t Transform origin [Default: false]
--timeout, -T <type> Timeout in milliseconds [Default: 60000]
--extension, -e <type> File extension [Default: webp]
--concurrency, -c <type> Concurrency level [Default: 3]
--quality, -q <type> Image quality [Default: 100]
--max-screenshots, -m <type> Maximum number of screenshots [Default: 0]
--recommended-size, -r <type> Recommended image size [Default: {"width":1200,"height":630}]
--output-dir, -o <type> Output directory [Default: ./public/screenshots]
--window-size, -w <type> Browser window size [Default: 1300,1300]
--chrome-path <type> Path to Chrome [Default: detected Chrome path]
--imagemagick-path <type> Path to ImageMagick [Default: detected ImageMagick path]
--overwrite Overwrite existing files [Default: false]. By default, the command will not take screenshots if the output file already exists.
Examples
$ og-screenshots --url "http://example.com"
Run
npx tsx source/cli.tsx --url "https://example.com"- Sometimes the process hangs. You can try to kill the process manually. Anyone who knows how to fix this issue is welcome to contribute.
Affero General Public License v3.0