diff --git a/MarkdownSource/docnetjson.md b/MarkdownSource/docnetjson.md
index a152c4c..48335cb 100644
--- a/MarkdownSource/docnetjson.md
+++ b/MarkdownSource/docnetjson.md
@@ -23,7 +23,8 @@ DocNet uses a json file to determine what to do in what form. The format is stra
             "__index" : "index.md",
             "Title Page 3" : "filename_page3.md",
             "Title Page 4" : "filename_page4.md"
-        }
+        },
+        "Auto generated" : "reference/**"
     }
 }
 ```
@@ -46,3 +47,14 @@ Each level, starting with the root, has a special page defined, `__index`. This
 
 ## Levels without __index defined
 If a level has no `__index` defined, `DocNet` will create a `__index` entry for the level and will specify as target `<path to index of parent>/nameoflevel.md`. If the page exists it will be loaded as the content for the index of the level, if it doesn't exist, the HTML will simply contain the topictitle and a list of all the sub topics in the level. This guarantees the tree can always be navigated in full.  
+
+## Use of wildcard inclusions
+If a level has a string value ending with `**`, it will process all .md files in the folders and subfolders and generate htm files from them. It will use the following process:
+
+1. Search (recursively) for all markdown (`.md`) files inside the specified folder (e.g. `reference`)
+2. Generate index (htm) files for all folders
+3. The title of the markdown files will be retrieved from the actual markdown files (so the first non-empty line will be used as title)
+
+@alert info
+The wildcard inclusion feature is extremely useful for referene or API documentation (mostly generated by a 3rd party tool)
+@end 
\ No newline at end of file