diff --git a/Exercises/__init__.py b/Exercises/__init__.py
deleted file mode 100644
index e69de29..0000000
diff --git a/Exercises/exercise-0/__init__.py b/Exercises/exercise-0/__init__.py
deleted file mode 100644
index e69de29..0000000
diff --git a/main.py b/main.py
index 51432a8..bccb934 100644
--- a/main.py
+++ b/main.py
@@ -1,13 +1,80 @@
 """
 This main entry point is the command line interface for the Speckle Automate function.
 """
+import random
 
+from pydantic import Field
 from speckle_automate import (
-    execute_automate_function,
+  execute_automate_function, AutomateBase, AutomationContext,
 )
 
-import Exercises.exercise_0.inputs.FunctionInputs as FunctionInputs
-import Exercises.exercise_0.function.automate_function as automate_function
+from Utilities.flatten import flatten_base
+
+
+class FunctionInputs(AutomateBase):
+    """These are function author defined values.
+
+    Automate will make sure to supply them matching the types specified here.
+    Please use the pydantic model schema to define your inputs:
+    https://docs.pydantic.dev/latest/usage/models/
+    """
+
+    comment_phrase: str = Field(
+      title="Comment Phrase",
+      description="This phrase will be added to a random model element.",
+    )
+
+
+def automate_function(
+  automate_context: AutomationContext,
+  function_inputs: FunctionInputs,
+) -> None:
+    """This is an example Speckle Automate function.
+
+      Args:
+          automate_context: A context helper object, that carries relevant information
+              about the runtime context of this function.
+              It gives access to the Speckle project data, that triggered this run.
+              It also has convenience methods attach result data to the Speckle model.
+          function_inputs: An instance object matching the defined schema.
+      """
+
+    # the context provides a convenient way, to receive the triggering version
+    version_root_object = automate_context.receive_version()
+
+    flat_list_of_objects = flatten_base(version_root_object)
+
+    # filter the list to only include objects that are displayable.
+    # this is a simple example, that checks if the object has a displayValue
+    displayable_objects = [
+      speckle_object
+      for speckle_object in flat_list_of_objects
+      if (
+           getattr(speckle_object, "displayValue", None)
+           or getattr(speckle_object, "@displayValue", None)
+         ) and getattr(speckle_object, "id", None) is not None
+    ]
+
+    if len(displayable_objects) == 0:
+        automate_context.mark_run_failed(
+          "Automation failed: No displayable objects found."
+        )
+
+    else:
+        # select a random object from the list
+        random_object = random.choice(displayable_objects)
+
+        automate_context.attach_info_to_objects(
+          category="Selected Object",
+          object_ids=[random_object.id],
+          message=function_inputs.comment_phrase,
+        )
+
+        automate_context.mark_run_success("Added a comment to a random object.")
+
+    # set the automation context view, to the original model / version view
+    automate_context.set_context_view()
+
 
 # make sure to call the function with the executor
 # Pass in the function reference with the inputs schema to the executor.