Rewrite and scraper rules

Both live in Feed Settings → Original Content.

Scraper rules

When the crawler is on (see Full article text), MicroRSS finds the article on the page by itself. A scraper rule overrides that with a CSS selector for the article's container, for example article .post-body or #content, .entry. Several selectors are comma-separated; their matches are joined in page order. The rule is used only while the crawler is on, and the crawler is a Pro feature.

Rewrite rules

Rewrite rules change the content after it is fetched, the feed's own content as much as a crawled article. The syntax is Miniflux's, functions separated by commas:

rule	what it does
remove(".share-bar, .newsletter")	drops every element matching the selector
replace("foo"|"bar")	replaces text, as a regular expression
add_dynamic_image	turns lazy-loaded images (data-src) into real ones
add_image_title	shows an image's title text under it
nl2br	turns line breaks into paragraphs, for feeds that send plain text
convert_text_link	makes bare addresses clickable
remove_clickbait	lowercases SHOUTED titles
remove_tables	unwraps layout tables

Any other Miniflux function is accepted and ignored, with a warning on save, so a rule set copied from Miniflux never breaks a feed.

The assistant

AI assistant in the same tab reads the article page and proposes a scraper rule and rewrite rules for it, which you can refine in a few words («keep the photo captions») and apply. It runs with your own key: Settings → AI Assistant takes the provider (Anthropic or OpenAI), the model and the key. The key stays in your browser and is sent only to the provider you chose; MicroRSS's server never sees it.