diff --git a/Introduction-to-basic-filtering-syntax.md b/Introduction-to-basic-filtering-syntax.md
new file mode 100644
index 0000000..8b4ad74
--- /dev/null
+++ b/Introduction-to-basic-filtering-syntax.md
@@ -0,0 +1,23 @@
+## Cosmetic filtering
+
+Example DOM tree:
+
+    <div id="unique-identifier" class="first-class second-class" data-storage="123-456">
+    <div class="first-class">
+
+- `##` - this is always present in cosmetic filters - this means this is cosmetic filter
+- `##div` - selecting by element type
+- `###unique-identifier` - selecting by `id`, `##` + `#` - this should be very efficient, because there should be only one unique identifier for page
+- `##.first-class` - selecting by `class`, `##` + `.` - will match other elements with same class, so:
+- `##.first-class.second-class` - use both classes to select only first line from example DOM.
+
+Using dot for classes automatically split value of the class attribute on space characters.
+
+Selecting by attribute:
+
+- `##[data-storage="123-456"]` - will match exactly
+- `##[data-storage^="123-"]` - will match from the beginning 
+- `##[data-storage$="-456"]` - will match from the end
+- `##[data-storage*="3-4"]` - will match in the middle
+- `##[id="unique-identifier"]` equivalent of `###unique-identifier`
+- `##[class="first-class"]` equivalent of `##.first-class`, but will match only second line from example DOM - class attribute is compared literally, no splitting on space characters.
\ No newline at end of file