https://extrastatic.dev/svekyll/svekyll-cli/-/merge_requests...
FWIW: I wrote a post using Svekyll about AI embeddings which has a view source button at the bottom. If you click that, you can download a full svekyll blog and add the RSS code to the _config.yml and then just "npm i && npm run build" to see it generate the RSS for that complex post. Then, look at it in your feed reader to see how it escapes the HTML, code blocks, images, etc.
Basically I write the posts in Markdown, commit them to the repository and it automatically generates a RSS feed and a index in the repository (inside the README.md file), and also publishes the blog posts to https://kokada.capivaras.dev/, that is where the blog actually lives.
I don't see how this is a problem, for proper preview, the markdown will have to get compiled into HTML/text anyway.
Most SSG's do this, but this can also be done with a good enough markdown compiler (cough lowdown) and a for loop. Without restricting your markdown formatting to a subset of the features of 2 markdown compilers
One question: how are the articles linked in the RSS file? Are they relative to that file and/or assumed to be in the same folder?
Does not currently support ordered or unordered lists.
